Skip to main content

Full text of "Psychology A Study Of A Science Volume 3 Formulations Of The Person And The Social Context"

See other formats


v.3        63-15383 
Psychology:  a  study  of 
a  science* 

K AN S AS  C IT Y  MO .  PU BLI C  LI :  B R A R V 

MM  MAY  1  A  1991 

Psycholgy:  A  Study  of  a  Science 


Volume  3.  Formulations  of  the  Person 
and  the  Social  Context 


The  Series 


Volume  1 .     Sensory,  Perceptual,  and  Physiological  Formulations 

CONTRIBUTORS:  Albert  A.  Blank,  James  J.  Gibson,  C.  H.  Graham,  D.  0.  Hebb, 
Harry  Hehon,  J.  C.  R.  Licklider,  Clifford  T.  Morgan,  Kenneth  N.  Ogle, 
M.  H.  Pirenne  and  F.  H.  C.  Marriott,  Leo  Postman  and  Edward  C.  Tolman, 
W.  C.  H.  Prentice 

Volume  2.     General    Systematic    Formulations,    Learning,    and    Special 


CONTRIBUTORS:  Dorwin  Cartwright,  Douglas  G.  Ellson,  W.  K.  Estes,  F.  C. 
Prick,  Edwin  R.  Guthrie,  Harry  F.  Harlow,  R.  A.  Hinde,  Arthur  L.  Irion, 
Frank  A.  Logan,  Neal  E.  Miller,  B.  F.  Skinner,  Edward  C.  Tolman 

Volume  3.    Formulations  of  the  Person  and  the  Social  Context 

CONTRIBUTORS:  Solomon  E.  Asch,  Raymond  B.  Cattell,  Franz  J.  Kallmann, 
Daniel  Katz  and  Ezra  Stotland,  Paul  F.  Lazarsfeld,  Henry  A.  Murray, 
Theodore  M.  Newcomb,  Talcott  Parsons,  David  Rapaport,  Carl  R.  Rogers, 
Herbert  A.  Thelen 


(These  titles  in  preparation) 

Volume  4.    Biologically  Oriented  Fields:  Their  Place  in  Psychology  and  in 
Biological  Science 

Volume  5.     The  Process  Areas,  the  Person,  and  Some  Applied  Fields:  Their 
Place  in  Psychology  and  in  Science 

Volume  6.     Investigations  of  Man  as  Socius:  Their  Place  in  Psychology  and 
the  Social  Sciences 


Volume  7.     Psychology  and  the  Human  Agent:  A  View  of  Problems  in  the 
Enaction  of  a  Science  (by  Sigmund  Koch] 

Psychology:  A  Study  of  a  Science 


Volume  3.  Formulations  of  the  Person 
and  the  Social  Context 

Edited  by  Sigmund  Koch 



New  York         Toronto         London 


PSYCHOLOGY:  A  STUDY  OF  A  SCIENCE  was  made  possible 
by  funds  granted  by  the  National  Science  Foundation  to  the  Amer- 
ican Psychological  Association,  and  carried  out  under  the  spon- 
sorship of  the  latter  organization.  Neither  agency,  however,  is  to 
be  construed  as  endorsing  any  of  the  published  findings  or  con- 
clusions of  the  Study, 

Copyright  ©  1959  by  the  McGraw-Hill  Book  Company,  Inc.  Printed 
in  the  United  States  of  America.  All  rights  reserved.  This  book, 
or  parts  thereof,  may  not  be  reproduced  in  any  form  without  permis- 
sion of  the  publishers.  Library  of  Congress  Catalog  Card  Number  57-14691 




When  one  looks  back  over  the  history  of  science,  the  successes  are 
likely  to  be  stressed  and  the  failures  forgotten.  Thus  one  tends  to  see 
science  as  starting  with  a  sure  sense  of  direction  and  progressing  neatly 
to  its  present  form.  Or  so  it  is  for  the  older  and  well  established 
branches  of  science;  but  not  for  psychology.  Psychology  has  not  one 
sure  sense  of  direction  but  several  quite  unsure  directions.  Growth  is 
erratic  and  there  is  much  casting  about  for  the  most  crucial  problems 
and  the  most  powerful  methods.  These  apparent  differences  between 
psychology  and  the  older  branches  of  science  may  result  from  the 
djfficul^^  it  is  perhaps  significant  that 

many  of  the  problems  of  psychology  were  not  attacked  by  the  methods 
of  science  until  so  late  a  date  in  history.  Or  the  differences  may  be  an 
illusion  resulting  from  the  much  closer  view  we  have  of  the  beginning 
struggles  to  develop  a  science  of  psychology  than  we  now  have  of  the 
beginning  efforts  in  the  older  sciences. 

Certainly  psychology  has  its  problems,  and  they  are  not  easy. 
Nevertheless,  knowledge  has  grown  rapidly  in  the  short  history  of 
man's  efforts  to  develop  a  science  of  behavior,  and  the  time  seems  ap- 
propriate for  a  major  effort  to  examine  the  progress  that  has  been 
made  in  attempting  to  find  a  way,  or  ways,  to  the  attainment  of  the 
explanatory  power  that  we  like  to  think  of  as  characteristic  of  science. 
A  growing  body  of  empirical  information,  a  serious  concern  over 
methodological  issues,  and  a  variety  of  efforts  to  bring  a  selected  body 
of  fact  into  the  organizing  framework  of  theory  all  emphasize  the  need 
for  that  line  of  questioning — always  going  on  in  science — which 
explores  the  shape  of  knowledge,  the  range  and  inner  connections  of 
the  ideas  through  which  it  has  been  developed  and  organized,  the 
changing  substructures  of  empirical  data,  and  their  emerging  relations 
to  each  other  and  to  the  findings  of  other  sciences.  The  seven  volumes 
of  Psychology:  A  Study  of  a  Science  are  a  response  to  this  need. 

The  first  three  volumes,  which  bear  the  collective  title  Study  L 
Conceptual  and  Systematic,  are  concerned  with  many  of  the  systematic 
formulations  of  recent  and  current  influence  which  psychologists  have 
developed  to  account  for  the  phenomena  in  which  they  are  interested. 


Each  systematic  position  is  analyzed  by  its  originator,  or  a  person  con 
nected  with  its  development,  in  a  way  which  gives  attention  to  th 
problems  it  seeks  to  solve,  the  empirical  basis  on  which  it  rests,  it 
degree  of  success,  and  its  relations  to  other  formulations. 

A  second  set  of  three  volumes,  collectively  called  Study  II.  Empirica 
Substructure  and  Relations  with  Other  Sciences,  inquires,  again  througl 
the  efforts  of  creatively  active  investigators,  into  the  organization  o 
various  fields  of  empirical  knowledge,  the  relations  of  one  to  another 
and  to  work  going  forward  in  other  sciences.  It  also  examines  suet 
problems  in  reverse  through  the  participation  of  social  and  biologicaJ 
scientists  who  consider  the  relations  of  their  own  special  fields  to  vari- 
ous parts  of  psychology.  The  three  volumes  of  Study  II,  now  in  prepa- 
ration, will  be  published  at  a  later  date. 

Volume  7 — Psychology  and  the  Human  Agent — will  present  the  Study 
Director's  view  of  certain  problems  of  psychological  inquiry  in  the  light 
of  the  findings  of  the  project. 

Primary  credit  for  the  initiation  of  these  studies  goes  to  the  Asso- 
ciation's Policy  and  Planning  Board,  which  decided  in  1952  that  the 
time  had  come  for  a  thorough  and  critical  examination  of  the  status 
and  development  of  psychology.  The  National  Science  Foundation 
agreed  upon  the  desirability  of  such  an  undertaking  and  has  gener- 
ously supported  the  effort.  When  funds  from  the  National  Science 
Foundation  were  found  to  be  insufficient  for  all  of  the  expenses  of  the 
studies,  the  American  Psychological  Association  provided  the  sup- 
plementary funds  necessary  to  complete  the  work. 

From  the  beginning,  the  study  was  divided  into  two  parts.  One  part 
dealt  with  the  education  of  psychologists  and  the  factors  conducive  to 
research  productivity  in  psychology.  That  part  was  directed  by  Profes- 
sor Kenneth  Clark  of  the  University  of  Minnesota,  who  has  reported 
the  findings  in  America's  Psychologists:  A  Survey  of  a  Growing  Profession, 
published  by  the  American  Psychological  Association  in  1957. 

The  other  part,  the  part  with  which  the  present  series  of  volumes  is 
concerned,  has  dealt  with  the  substance  of  psychological  thought  and 
data.  Professor  Siginund  Koch  of  Duke  University  has  been  responsible 
for  this  part  of  the  study.  Working  closely  with  him  has  been  a  panel  of 
consultants  consisting  of  Lyle  H.  Lanier,  Howard  H.  Kendler,  Conrad 
G.  Mueller,  and  Karl  E.  Zener.  These  men,  but  chiefly  Dr.  Koch,  have 
planned,  organized,  interpreted  and  edited  the  work,  and  successfully 
enlisted  the  cooperation  of  the  approximately  80  authors  whose  origi- 
nal papers  will  constitute  the  basic  material  of  the  series. 

In  the  background,  at  a  safe  distance  from  the  labors  that  have 
sometimes  engulfed  Dr.  Koch,  his  panel  of  consultants,  and  the  pri- 
mary authors,  has  been  a  steering  committee  on  which  I  had  the  pleas- 

Preface  vii 

ure  of  serving  as  chairman,  and  having  as  colleagues  Clarence  H. 
Graham,  Lyle  H.  Lanier,  Robert  B.  MacLeod,  Eliot  H.  Rodnick,  M. 
Brewster  Smith,  and  Robert  L.  Thorndike.  The  steering  committee 
helped  to  make  administrative  arrangements  and  helped  to  decide 
on  the  scope  of  the  studies,  but  takes  no  credit  for  their  successful 

In  the  preface  to  Americans  Psychologists  we  have  already  acknowl- 
edged our  gratitude  to  Kenneth  Clark  and  his  collaborators  who 
helped  to  produce  that  volume.  It  is  our  final  pleasant  duty  to  express 
our  thanks  to  Duke  University  for  making  Dr.  Koch's  time  available; 
to  the  National  Science  Foundation  for  its  necessary  and  generous 
financial  support  and  for  the  counsel  and  support  of  John  T.  Wilson, 
Assistant  Director  for  the  Biological  Sciences;  to  Lyle  H.  Lanier, 
Howard  H.  Kendler,  Conrad  G.  Mueller,  and  Karl  E.  Zener  for  their 
critical  and  devoted  help;  to  all  of  the  authors  whose  names  appear  on 
the  title  pages  for  their  original  contributions;  and — most  of  all — to 
Sigmund  Koch  for  directing  and  driving  through  to  completion  what 
we  hope  will  be  an  oft-consulted  aid  to  the  scholars  and  research 
workers  who  are  striving  to  increase  the  rigor  and  further  the  develop- 
ment of  scientific  psychology. 

Dael  Wolfle,  CHAIRMAN 




Preface v 

Dael  Wolfle 

Introduction  to  Volume  3 1 

Sigmund  Koch 

Preparations  for  the  Scaffold  of  a  Comprehensive  System     .      .         7 

Henry  A.  Murray 

The  Structure  of  Psychoanalytic  Theory:  A  Systematizing  At- 
tempt     55 

David  Rapaport 

A  Theory  of  Therapy,  Personality,  and  Interpersonal  Relation- 
ships, as  Developed  in  the  Client-centered  Framework     ,      184 

Carl  R.  Rogers 

Personality  Theory  Growing  from  Multivariate  Quantitative  Re- 
search     257 

Raymond  B.  Cattell 

Psychogenetic  Studies  of  Twins 328 

Franz  J.  Kallmann 

A  Perspective  on  Social  Psychology 363 

Solomon  E.  Asch 

Individual  Systems  of  Orientation 384 

Theodore  M.  Newcomb 

A  Preliminary  Statement  to  a  Theory  of  Attitude  Structure  and 

Change 423 

Daniel  Katz  and  Ezra  Stotland 

Latent  Structure  Analysis 476 

Paul  F.  Lazarsfeld 



Work-emotionality  Theory  of  the  Group  as  Organism     .      .      .     544 

Herbert  A.  Thelen 

An  Approach  to  Psychological  Theory  in  Terms  of  the  Theory  of 

Action 612 

Talcott  Parsons 


Suggested  Discussion  Topics 713 

Note  on  the  Use  of  Discussion  Topic  Index  Numbers  .      .      .      724 

Some  Trends  of  Study  I  (Vols.  1-3) 

Epilogue 729 

Sigmund  Koch 
Significance  of  Sensory  Psychology  for  Certain  Methodological 

Problems 789 

Conrad  G.  Mueller 

Name  Index 803 

Subject  Index 813 


Psychology:  A  Study  of  a  Science  is  a  report  of  inquiries  into  the  status 
and  tendency  of  psychological  science.  Some  eighty  distinguished  authors 
have  contributed  sustained  essays  which  consider:  (Study  I)  major 
theoretical  formulations  of  recent  importance;  and  (Study  II)  the  struc- 
ture, mutual  interrelations,  and  associations  with  other  sciences  of  the 
main  empirical  areas  in  which  psychological  research  is  pursued.  The 
findings  of  Study  I  Conceptual  and  Systematic  comprise  the  initial 
three  volumes  of  the  series;  Study  II  Empirical  Substructure  and  Rela- 
tions with  Other  Sciences  is  reported  in  the  following  three  volumes.  A 
final  volume  by  the  Study  Director — Psychology  and  the  Human  Agent 
— includes  commentary  on  the  significance  of  the  findings. 

The  present  volume  is  the  third  in  the  series  and  is  part  of  Study  I. 
Each  of  the  eleven  essays  in  this  book  is  a  self-contained  presentation 
which  may  be  read  with  profit  independently  of  the  others,  or  of  the  con- 
tents of  other  volumes.  Yet  the  reader  must  bear  in  mind  that  the  present 
volume  is  the  receptacle  of  a  fragment  of  Study  I,  and  that  Study  I,  in 
turn,  is  part  of  a  larger  enterprise  having  certain  unifying  values,  aims, 
and  methods.  For  a  conception  of  these  latter,  the  reader  is  referred  to 
Dael  Wolfle's  Preface  and  to  the  two  detailed  introductory  sections  in 
Volume  1 — one  for  the  series,  the  other  specifically  for  Study  I. 

As  an  immediate  aid  to  the  reader,  a  resume  of  the  Study's  design  is 

Study  I  Conceptual  and  Systematic.  This  study  involved  the  in- 
tensive analysis  of  thirty-four  "systematic  formulations"  of  widely  vary- 
ing type  and  subject-matter  reference  and  all  of  established  influence 
in  recent  psychology.  A  "systematic  formulation"  was  defined  quite 
generally  as  "any  set  of  sentences  formulated  as  a  tool  for  ordering 
knowledge  with  respect  to  some  specified  domain  of  events,  or  further- 
ing the  discovery  of  such  knowledge" :  in  applying  this  definition,  care 
was  taken  that  no  formulation  be  precluded  by  nonconformity  to  stand- 
ardized conceptions  of  the  nature  of  "theory."  Since  each  systematic 
formulation  is  the  end-product  of  a  human  effort  to  see  and  state  order 
in  a  given  domain,  each  analysis  was  made  either  by  the  originator  (s) 



of  the  formulation  in  question  or  (in  a  few  cases)  by  individuals  crea- 
tively associated  with  the  development  of  formulations  of  which  they 
were  not  the  primary  authors. 

Each  systeniatist  was  invited  to  approach  his  work  with  certain 
common  themes  of  analysis  in  mind.  These  were  designed  to  invite  a 
convergence  of  insight  on  those  problems  of  systematization  which  had 
emerged  from  the  practice  of  the  past  three  decades,  more  or  less.  Some 
of  the  suggested  problems  had  been  conspicuous  in  previous  "meta- 
systematic"  discussion,  but  required  in  our  opinion  exposure  to  a  wider 
range  of  systematically  schooled  sensibilities.  Others  were  problems  that 
seemed  critically  posed  by  recent  systematic  work,  yet  ones  which  had 
received  little  or  no  explicit  attention. 

The  dominating  hope  was  for  analyses  that  might  illumine  the 
relations  between  the  creative  processes  of  systematizing  and  their  publicly 
expressed  products.  It  was  thus  hoped  that  the  atmosphere  of  the  study 
might  encourage  as  much  concern  with  background  influences,  orienting 
presuppositions,  and  working  methods,  as  with  conceptual  content,  re- 
search achievements,  and  prospects.  It  was  felt  that  analysis  of  this  order 
could  itself  have  creative  consequences;  reflective  scrutiny  of  the  extent 
and  depth  envisaged  means  rethinking.  The  primary  intent  of  the  dis- 
cussion themes  (and  indeed  the  constant  aim  of  all  editorial  effort)  was 
to  realize  an  atmosphere  that  might  invite  such  emphases.  Authors  were 
requested  to  make  explicit  reference  to  the  themes  in  their  writing  only 
to  an  extent  they  deemed  appropriate  or  congenial.  The  use  of  the 
themes  for  facilitating  the  collation  of  findings  was  thus  a  secondary,  if 
still  important,  aim.  As  matters  turned  out,  most  authors  adhered  to  them 
sufficiently  to  give  the  reader  an  excellent  purchase  for  the  detection  of 
similarities  and  differences  on  key  issues. 

The  grounds  for  the  selection  of  the  thirty-four  formulations  in- 
cluded in  Study  I  are  given  in  Volume  1  (pp.  21-27).  The  aim  was 
a  reasonably  balanced  diversification  of  formulations  (as  judged  by 
many  consultants)  with  respect  to  (a]  subject-matter  reference,  and  (b) 
conceptual  and  methodological  "type,"  Many  significant  formulations 
that  we  would  have  wished  to  represent  in  the  original  list  were  excluded 
by  spatial  and  other  arbitrary  restrictions.  Nor  are  all  formulations 
originally  chosen  included  in  the  present  volumes.  Though  the  proportion 
of  inclusions  is  remarkable,  there  were  some  individuals  who  could  not 
participate.  We  do  not,  then,  claim  "representativeness"  even  in  an  in- 
formal and  impressionistic  sense.  We  do,  however,  claim  sufficient  di- 
versity to  extend  markedly  the  range  of  formulations  which  in  recent 
years  have  been  given  sustained  analytic  attention. 

Study  II  Empirical  Substructure  and  Relations  with  Other 
Sciences.  This  study  seeks  increased  understanding  of  the  internal 

Introduction  to  Volume  3  3 

structure  of  psychological  science  and  its  place  in  the  matrix  of  scientific 
activity.  Over  forty  contributors,  having  distinguished  research  back- 
grounds in  psychology,  or  in  related  biological  and  social  sciences,  were 
invited  to  write  papers  which  examine  the  organization  of  empirical 
knowledge  within  subareas  of  these  disciplines,  and  which  chart  their 
cross  connections.  Psychologist  contributors  consider  the  relations  be- 
tween their  own  fields  of  special  competence  and  the  rest  of  psychology, 
and  inquire  also  into  relations  with  relevant  segments  of  other  sciences. 
Social  and  biological  science  contributors  examine  the  relations  between 
their  own  fields  and  psychology. 

All  authors  are  individuals  whose  research  interests  have  bridged 
conventionally  discriminated  fields  of  knowledge.  Each  was  asked  to 
place  special  emphasis  on  those  "bridging  problems"  which  had  been 
central  in  his  own  research  experience.  As  in  the  case  of  Study  I,  an 
attempt  was  made  to  encourage  differentiated  and  stratified  analysis, 
and  to  invite  a  convergence  of  visions  on  significant  cross-cutting  issues, 
by  proposing  certain  common  themes  of  analysis.  The  "themes"  for 
Study  II  comprise  a  detailed  breakdown  of  the  senses  in  which  questions 
of  "mapping"  subject-matter  structure  and  exploring  field  interrelations 
might  be  entertained. 

Though  the  topography  of  a  science  is  too  vast  and  labile  for  com- 
prehensive or  final  mapping,  this  very  fact  makes  it  more  important 
to  assay  the  contours  of  knowledge  as  best  we  can.  Study  II  exploits 
the  only  resource  available  for  such  problems — individual  vision — but 
in  a  novel  way.  It  assumes  that  a  pooling  of  expert,  specialized  sensibilities 
can  give  insight  into  the  emerging  structure  of  a  science  of  a  sort  not 
ordinarily  available. 

A  fuller  statement  of  the  plan  for  Study  II  appears  in  the  General 
Introduction  to  the  Series  (Vol.  1,  pp.  1-18).  An  adequate  account 
of  working  methods  must  await  publication  of  the  completed  study. 

Psychology  and  the  Human  Agent.  This  volume  is  a  postscript  to 
the  Study,  representing  certain  views  formed  by  the  Director  in  its 
course.  The  book  (a)  records  those  attitudes  towards  a  science  and 
science  which  necessarily  color  the  spirit  of  the  Study,  (6)  constructs 
trends  from  the  massive  findings  of  the  two  group  studies,  and  (c) 
considers,  in  the  light  of  the  Study's  premises  and  apparent  trends, 
certain  problems  of  psychological  inquiry  suggested  by  the  practice  of 
the  past  several  decades. 

In  this  day  in  which  "self  studies"  and  reductions  of  enigma  by 
seminar  are  becoming  commonplace  in  social  science,  it  may  be  helpful 
to  mention  a  few  of  the  special  features  of  Psychology:  A  Study  of  a 
Science.  May  we  stress  (in  random  order)  the  following  points: 


1.  Both  group  studies  are  "collaborative"  but  only  in  the  special 
sense  that  many  creative  men  agreed  to  pursue,  within  the  climate  of 
the  Study,  individual  tasks  of  vital  interest  to  themselves.  The  study  is 
not  collaborative  or  "groupish"  in  any  sense  implying  an  intention  to 
relinquish  individuality  or  even  idiosyncrasy  to  some  prissy  conception  of 
the  common  weal. 

2.  The  Study  aims  for  no  grand  "integration"  of  knowledge.  If  a 
"Summa  Psychologica"  or  even  a  "Synopticon"  had  seemed  even  re- 
motely within  reach,  our  inclination  would  have  been  to  abstain  on 
principle.  The  Study  seeks  to  reflect  the  diversity  of  thought  that  actually 
exists  and  is  premised  on  th^jv^ 

proachjas^^  It  conceives  of  its  con- 

tributor groups  as  pluralities  of  creative  individuals  who  view  those  areas 
which  they  know  best  through  the  screen  of  their  own  expert  sensibilities. 
The  discussion  outlines  invite  the  play  of  individual  sensibilities  on  com- 
mensurable themes,  thus  helping  the  reader  to  collate  positions. 

3.  The  Study's  aims  are  neither  legislative  nor  evaluative.  It  rejects 
all  monolithic  codes  for  the  generation  and  processing  of  knowledge,  or 
for  virtuous  scientific  conduct.  If  there  is  a  central  bias,  it  is  for  the 
loosening  of  those  constraints  which  can  keep  men  from  significant  prob- 
lems or  thoughts  through  fear  of  the  unorthodox. 

The  Study  by  no  means  devalues  the  insights  of  recent  "science 
of  science"  but  would  wish  them  set  in  a  perspective  better  adjusted 
to  a  field  barely  beginning  to  test  established  methods  of  science  on  an 
inimitably  diverse  and  intricate  collection  of  subject  matters.  Such  a 
perspective  can  emerge  by  seeing  the  end-products  of  science  as  every- 
where conditioned  by  human  decision,  value,  creative  option ;  by  freeing 
from  staleness  that  truism  which  holds  scientific  inquiry  to  be  con- 
tinuous with  other  human  activities.  Analysis  in  science,  then,  becomes^ 
more  than  a  succession  of  routine  tasks  in  the  "logic"  of  science;  it  be- 
comes an  enterprise  which  can  uncover  the  significance  of  its  objects 
only  by  holding  in  view  the  relations  between  creative  process  and 
sentential  product,  strategic  gamble  and  cognitive  outcome.  *"* 

The  grounds  for  the  inclusion  of  the  eleven  formulations  contained 
in  this  volume  are  best  conveyed  by  reference  to  the  planning  for  the 
total  Study  (cf.  Introduction  to  Study  I,  Vol.  1,  especiaUy  pp.  21-27). 
Eleven  topics  is  a  stingy  allotment  relative  to  the  range  and  density  of 
effort  in  studies  of  the  person  and  the  social  setting.  If  many  significant 
lines  of  work  have  been  omitted,  the  ones  included  are  also  significant— 
and  sufficiently  varied  to  suggest  a  generous  range  of  the  problems  and 
tasks  that  systernatists  face. 

Of  the   contributions  relevant  to    "personality,"   three— Murray's 

Introduction  to  Volume  3  5 

"scaffold,"  Rapaport's  systematic  examination  of  psychoanalysis,  and 
Rogers's  codification  of  his  client-centered  framework — are  representative 
of  conceptual  frameworks  having  broad  scientific  objectives  and  long- 
standing influence.  Cattell  draws  together  certain  proposals  towards 
systematic  thought  suggested  by  the  logic  and  findings  of  one  of  the 
principal  methods  of  personality  research  (factor  analysis).  By  reviewing 
a  delimited  but  basic  problem  area  (psychogenetic  studies  of  twins), 
Kallmann  gives  an  exhibit  of  the  type  of  painstaking,  stepwise  empirical 
work  on  which  the  resolution  of  issues  common  to  many  systematists 
must  depend. 

Turning  to  the  contributions  that  most  would  allocate  to  "social 
psychology,"  again  we  find  represented  three  lines  of  work  which  point 
towards  general  systematic  objectives.  Each  of  these,  moreover,  stresses 
a  different  one  of  three  principal  "levels53  at  which  social  "variables" 
may  be  conceived.  Asch  could  be  said  to  represent  the  type  of  social 
psychology  which  sees  no  basis  for  conceptual  separation  from  individual 
psychology.  Thelen  and  Parsons  consider  instances  of  the  approaches  of 
"group  dynamics"  and  "unified  social  science  theory,"  respectively.  An 
important  methodic  formulation  (latent  structure  analysis)  is  offered  by 
Lazarsfeld  in  response  to  a  class  of  problems  which  he  sees  as  ubiquitous 
to  psychology  and  social  science.  Finally,  the  contributions  of  Newcomb 
and  of  Katz  and  Stotland  present  preliminary  systematic  assessments  of 
problems  stemming  from  one  of  the  more  active  fields  of  social  psy- 
chological research — the  study  of  attitudes. 

On  one  point,  agreement  among  authors  is  so  vehement  that  it 
may  be  not  unfair  to  anticipate  it  here.  Personologists  and  psychologists 
called  "social"  refuse  in  this  volume  to  compress  their  concerns  into  the 
standard  compartments.  If  they  have  always  seen  their  concerns  as  funda- 
mental to  the  task  of  psychology,  they  now  seem  eager  to  assert  this  even 
within  the  hearing  of  "fundamental"  psychologists.  To  such  refreshing 
truculence  we  could  but  yield  by  avoiding  the  time-worn  substantives 
"personality"  and  "social  psychology"  in  the  title  of  this  volume. 

Mention  of  certain  editorial  provisions  is  in  order.  Readers  will  find 
the  complete  statement  of  discussion  themes,  as  sent  to  contributors, 
reproduced  in  the  Appendix.  There  is,  of  course,  variation  in  the  extent 
to  which  the  different  presentations  adhere  to  the  discussion  themes. 
As  an  aid  to  readers  interested  in  the  detection  of  key  convergences  and 
divergences  of  positions,  index  numbers  corresponding  to  the  principal 
thematic  items  have  been  inserted,  where  relevant,  in  the  individual 
tables  of  contents  appearing  before  each  paper.  The  system  of  index 
numbers  is  explained  in  the  Appendix. 

This  final  volume  of  Study  I  contains  a  section  of  general  comment 
on  the  study.  An  editorial  epilogue  presents  a  few  trends  suggested  by 


the  essays  in  all  three  volumes.  Attention  is  restricted  to  conspicuous 
trends  which  can  give  a  "fix"  on  the  position  of  systematic  psychology 
relative  to  its  recent  history.  There  is  also  a  special  supplement  by  Conrad 
Mueller  on  certain  methodological  implications  of  the  contributions  in 
sensory  psychology.  Dr.  Mueller  generously  served  in  a  capacity  much 
like  that  of  sub-editor  in  the  sensory  area. 

Further  discussion  of  trends  will  be  offered  in  the  final  volume  of 
the  series,  Psychology  and  the  Human  Agent.  There  is,  however,  no 
standard  "theory53  of  the  meaning  of  this  study.  If  there  were,  we 
would  consider  the  main  aim  compromised.  That  aim  has  been  to 
develop  materials  of  unique  comprehensiveness  and  depth  in  terms  of 
which  each  reader  may  enrich  his  own  view  of  systematic  psychology. 
The  thirty-four  essays  of  Study  I  can  reward  efforts  towards  secondary 
analysis  and  synthesis — whether  by  student,  specialized  scholar,  or  gen- 
eral reader — for  a  long  time  to  come.  Let  there  be  as  many  theories  of 
this  study  as  there  are  readers. 

Psychology:  A  Study  of  a  Science  is  the  result  of  a  project  sponsored 
by  the  American  Psychological  Association  and  subsidized  by  the  Na- 
tional Science  Foundation.  The  project  was  known  as  "Project  A"  of  the 
"APA  Study  of  the  Status  and  Development  of  Psychology."  The  work 
profited  from  the  counsel  of  an  Advisory  Committee  consisting  of 
Dael  Wolfle,  Chairman,  and  Clarence  H.  Graham,  Lyle  H.  Lanier, 
Robert  B.  MacLeod,  Eliot  H.  Rodnick,  M.  Brewster  Smith,  and  Robert 
L.  Thorndike.  Howard  H.  Kendler,  Lyle  H.  Lanier,  Conrad  G. 
Mueller,  and  Karl  E.  Zener  composed  a  Panel  of  Consultants  to  the 
Director.  The  generous  part  played  by  the  members  of  both  groups  is 
described  in  the  introductory  sections  of  Volume  1. 


Harvard  University 

Introduction 7 

Interest  in  Significant  Human  Feelings,  Thoughts,  and  Actions.  Influence  of 

Medicine 9 

Interest  in  the  Earliest  and  Innermost  Origins  of  Things 12 

Interest  in  Process,  Change,  Differential  Development,  Creativity.  Influence  of 

Chemical  Embryology 14 

Observations  of  the  Interdependence  and  Hierarchical  Integration  of  Functional 

Processes:  Adoption  of  the  Organismic  Concept.  Influence  of  L.  J.  Henderson  17 

Interest  in  the  Directionalities  and  Effects  of  Overt  Behaviors 19 

Influence  of  Whitehead  and  Lewin:  Concepts  of  Physical  Field,  Cathexis,  Pro- 
ceeding, Serial,  etc 21 

Additional  concepts  for  the  present  scaffold 29 

Cathexis 29 

Dyadic  system 30 

Thema 31 

Thematic  dispositions 34 

Serials 34 

Ordination 35 

Influence  of  Freud,  Jung,  and  Other  Psychoanalysts 36 

Influence  of  Darwin,  Bergson,  and  Other  Evolutionists:  Adoption  of  the  Con- 
cept of  Creativity 38 

Influence  of  Social  Evolutionists,  Cultural  Anthropologists,  and  Sociologists     .  45 

Compelling  Need  for  Comprehensiveness 47 

Tolerance  of  Uncertainty 49 

Interest  in  Systems 50 

Apologia 52 

References 53 


It  seems  that  the  majority  of  my  voices  are  in  favor  of  this  enterprise, 
for  here  I  am,  pen  in  hand,  intending  to  comply  so  far  as  possible  with 
the  editorial  suggestions. 


8  HENRY   A.    MURRAY 

But  a  minority  of  me — and  now  surely,  at  the  outset,  is  the  moment 
to  give  vent  to  it — believes  that  certain  of  the  analyses  invited  by  the 
discussion  outline  are  premature,  not  for  all  psychologists  perhaps,  but 
for  those  who  are  concerned  with  human  lives  and  personalities.  The 
topics  suggested  for  discursive  treatment  are  broadly  defined;  but, 
even  when  taken  with  a  grain  of  salt,  the  task  calls  for  meticulous 
criticism  of  one's  own  speech,  semantic  niceties,  overelegant  definitions. 
Should  not  criticism  and  refinement  be  in  balance  with  spontaneity,  ex- 
ploration, and  invention  if  a  science  is  to  grow  in  a  way  and  at  a  pace 
appropriate  to  its  age?  Also,  do  we  have  sufficient  data  or  sufficient 
organization  of  the  data  to  arrive  at  anything  more  than  a  miniature 
system  for  a  tiny  region  of  transactions?  Systematic  psychology,  being 
very  young,  has  occupied  only  a  small  portion  of  its  legitimate  terrain. 
Its  contemporary  schools  are  like  our  thirteen  colonies  along  the  Atlantic 
coast  line,  a  narrow  strip  of  provincial  culture.  Their  manifest  destiny 
is  to  move  West,  order  the  wilderness  with  the  best  available  tools,  crude 
as  they  now  are,  and  eventually  achieve  a  more  refined  and  compre- 
hensive system  which  embraces  all  parts  and  functions  of  the  whole,  the 
total  personality.  At  this  stage  I  should  hate  to  see  our  center  of  gravity 
move  any  further  to  the  side  of  perfectionistic  rituals,  a  hair-splitting  fussy 

No  doubt  this  large  endeavor  will  bear  fruit;  but  despite  its  promise, 
it  is  not  applauded  at  this  moment  by  some  members  of  my  household 
because  of  their  suspicion  that  it  is  liable  to  seduce  some  promising 
psychologists  away  from  the  study  of  personalities — the  domain  that  is 
theirs,  and  only  theirs,  to  explore,  survey,  and  map — away  from  the 
humanistically  important  riddles  which  we  should  be  creeping  up  on 
gradually  and  craftily. 

Another  reason  for  my  hesitation  in  joining  this  enterprise  is  the 
impossibility  of  my  adhering  to  the  suggested  ordinance  of  discourse.  It 
is  evident  that  certain  of  its  terms  could  be  met  only  by  psychologists 
with  other  aims  than  mine.  It  is  an  admirable  mold — straightedged  and 
nicely  shaped — for  exclusively  experimental  specialists,  observers  of 
closely  restricted  animal  activities,  peripheralists,  and  positivists;  but 
literal  adherence  at  all  points  is  scarcely  possible  for  naturalists,  gen- 
eralists,  and  centralists,  who  study  gradual  transformations  of  the  dis- 
positions, beliefs,  and  modes  of  action  of  human  beings  as  they  manifest 
themselves  in  different  social  settings. 

Despite  the  above  reservations,  twenty  months  ago  it  was  decided 
somehow  that  I  accept  the  challenge  as  an  adventure  in  self-discipline; 
and,  in  conformity  with  the  committee's  outline,  I  went  ahead  with 
what  amounts  to  an  intellectual  autobiography  in  so  far  as  this  relates 
to  the  development  of  my  present  scaffold  for  a  theory  of  personality. 

Preparations  for  Scaffold  of  a  Comprehensive  System        9 

This  part  of  the  assignment  was  easier  than  I  anticipated;  but  the  second 
part — setting  up  a  logically  articulated  skeleton  of  the  whole — was  so 
much  more  difficult  that,  despite  an  extension  of  time  as  well  as  every 
possible  guidance  and  encouragement  from  a  most  charitable  Director,  I 
was  unable  to  arrive  at  a  satisfactory  set  of  basic  propositions  before 
the  date  line.  In  short,  I  proved  unequal  to  the  set  standard.  It  happened, 
however,  that  more  than  half  of  the  matter  to  be  ordered  in  Part  2  has 
been  included  in  Part  1,  and  so,  the  Director,  pressed  by  generosity, 
decided  that  the  peculiar  fragment  which  lies  before  you  might  serve  as  a 
kind  of  substitute  contribution.  Its  title  might  be  this:  certain  orienting 
dispositions,  impressive  observations,  and  influential  theories  as  deter- 
minants of  scientific  aims,  assumptions,  methods,  and  conceptions. 


It  is  generally  assumed  by  the  uninformed  and  innocent  that  all 
psychologists  must  have  at  least  one  "orienting  attitude55  in  common:  a 
stout  affection  for  human  beings  coupled  with  a  consuming  interest  in 
their  emotions  and  evaluations,  their  imaginations  and  beliefs,  their 
purposes  and  plans,  their  endeavors,  failures,  and  achievements.  But 
this  assumption,  it  appears,  is  not  correct.  A  psychologist  who  has  been 
constantly  prodded  and  goaded  by  these  propulsions,  as  I  have  been, 
belongs  to  a  once  small  and  feeble,  though  now  expanding  and  more 
capable,  minority.  Anyhow,  this  bent  of  empathy  and  curiosity  toward 
all  profound  experiences  of  individual  men  and  women  should  be  set 
down  as  one  of  the  prime  determinants  of  several  definitive  decisions, 
which  shall  be  mentioned,  respecting  the  scope  of  my  scientific  concern 
and  of  a  methodology  to  fit.  This  is  a  crucial  point  because,  if  my  interest 
in  events  of  this  sort  had  been  less  steadfast,  I  might  have  turned  to  more 
manageable  phenomena. 

My  interest  in  people,  their  doings  and  their  ills,  must  have  had 
something  to  do  with  my  choice  at  college  of  history  as  field  of  con- 
centration and  of  medicine  as  career  for  later  life. 

The  study  of  history  implanted  the  idea  of  the  time  dimension  as  an 
essential  part  of  the  very  definition  of  reality  as  well  as  a  miscellany  of 
coarse  facts  to  support  my  speculations  when  I  dipped,  three  decades 
later,  in  the  stream  of  sociology  and  anthropology.  But  the  study  of 
medicine  was  more  influential:  it  led  to  two  years  of  surgery  and  five 
years  of  research  in  physiology  and  in  the  chemistry  of  embryology,  with 
a  Ph.D.  from  Cambridge  University  in  physiological  chemistry. 

The  practice  of  medicine  taught  me  a  lot  of  commonsensical  things, 
one  of  which  was  that  among  the  few  almost  indispensable  methods  of 

10  HENRY   A.    MURRAY 

arriving  at  valid  diagnoses  (apperceptions,  inferences)  is  that  of  inquiry 
— the  thorough  detailed  recording  of  the  patient's  memories  of  interior 
sensations  and  pertinent  emotional  experiences.  We  were  taught  to 
distinguish  perceptible  physical  signs  (overt  sense  data)  and  imper- 
ceptible symptoms  (reports  of  covert  psychic  processes)  and  to  value 
both.  The  proof  obtained  on  the  operating  table,  time  and  time  again, 
that  a  correct  diagnosis  of  an  abdominal  condition  could  be  made 
solely  on  the  basis  of  a  patient's  reported  symptoms  was  so  firmly  im- 
printed on  the  entablatures  of  my  cortex  that  when,  in  later  years,  I  was 
confronted  by  Watson's  dogma — his  radical  repudiation  of  subjective  ex- 
periences as  material  for  psychology — my  head  assigned  it  to  the  category 
of  eccentric  foibles.  I  was  an  empirical  behaviorist,  born,  bred,  and 
trained,  in  the  sense  that  every  physicist,  chemist,  and  biologist  is  neces- 
sarily a  behaviorist.  But  when  it  came  to  dealing  with  human  beings,  I 
could  see  no  advantage  in  allowing  myself  to  be  converted  into  an 
exclusive,  half-paralyzed  behaviorist  who,  on  metaphysical  grounds,  elects 
to  deny  himself  an  invaluable  source  of  data.  (This  does  not  apply  to  the 
current  cultural  situation:  today,  after  a  complete  semantic  somersault, 
every  psychological  process — perception,  emotion,  dreaming — is  called 

My  above-mentioned  interest  in  people  was  not  at  all  confined  to 
their  physical  activities — say,  to  the  routes  they  chose  and  the  muscles 
they  used  in  locomoting  to  the  restaurants  they  preferred  to  ingest  the 
food  that  was  most  appealing  to  their  senses.  I  was  much  more  interested 
in  their  feelings,  evaluations,  and  conceptions  relative  to  other  matters, 
and  for  the  most  part,  so  were  they — and  so  were  my  militantly  be- 
havioristic  friends  of  later  years — more  interested  in  the  valued  products 
of  their  intellections  than  in  their  own  muscular  accomplishments.  In 
due  couree,  assured  that  correctness  of  prediction  is  the  best  index  of  the 
relative  worth  of  different  methods,  I  did  a  few  impromptu  experiments 
and  found  empirically  that  the  most  dependable  single  operation  I 
could  perform  in  attempting  to  foretell  what  a  behaviorist  would  do  next 
or  in  the  near  future  was  to  ask  him.  But  the  commonsensical  avowal  I 
wish  to  make  here  is  this:  that  first  as  a  doctor  and  second  as  a  psy- 
chologist I  have  never  ceased  to  elicit  direct  expressions  and  reports  of 
interior  experiences — somatic,  emotional,  and  intellectual — not  only  as 
sources  of  indications  of  overt  actions  to  be  executed  in  the  future,  but 
as  indications  of  occurrences  that  are  intrinsically  important.  For  ex- 
ample, the  occurrence  of  anxiety,  or  the  persistence  of  unhappiness, 
or  the  generation  of  a  new  theory  is  as  important  to  me  when  taken  as  a 
dependent  variable  (something  to  be  predicted)  as  it  is  when  taken  as 
a  hypothetical  or  intervening  variable  (an  aid  in  the  prediction  of  some- 
thing else).  Though  imperceptible  to  us  and  therefore  inferential.,  covert 

Preparations  for  Scaffold  of  a  Comprehensive  System      1 1 

mental  processes  and  products,  some  with  and  some  without  the  property 
of  consciousness,  happen  to  be  intrinsically  attractive  to  a  cogitator  of 
my  persuasion,  and  I  see  no  insuperable  barrier  to  their  being  incor- 
porated in  a  unified  body  of  scientific  facts  and  propositions.  If  the 
heavenly  bodies  had  memories  reaching  back  to  the  Big  Bang  and  words 
intelligible  to  us,  what  astronomer  would  shut  his  ears  to  them  on 
principle?  Anyhow,  this  concern  of  mine,  this  reliance  on  a  multiplicity 
of  inferences,  checked  and  rechecked,  this  vision  of  a  theoretical  system 
largely  composed  of  psychological,  rather  than  physical,  variables,  makes 
it  necessary  for  me  to  leave  exclusive  positivism  to  those  who  deal  with 
entities  that  are  incapable  of  supplying  us  with  valuable  verbal  repre- 
sentations of  what  has  occurred  and  is  occurring  behind  their  surfaces. 
But  enough  said;  I  must  return  to  my  surgical  internship  and  finish 
listing  what  I  learned  that  influenced  subsequent  decisions  respecting 
procedures  and  objectives  in  the  field  of  psychology. 

From  medical  practice  I  derived  the  "multiform  method"  of  assess- 
ment, coupled  with  the  belief  that  it  should  be  possible  for  a  group  of 
trained  collaborators  using  a  wide  variety  of  methods  to  make  a  reason- 
ably complete  examination,  formulation,  and  appraisal  of  a  whole  person 
as  an  ongoing  order  of  differentiated  functional  activities.  This  objective 
is  achieved  over  and  over  again  on  the  physiological  level  by  practitioners 
of  medicine  but  when  transferred  to  the  psychological  level  its  attain- 
ment is  impeded  by  innumerable  special  difficulties.  To  cut  down  my 
hope  to  size — to  make  it  congruent  with  what  can  feasibly  be  undertaken 
under  existing  conditions,  with  available  personnel,  with  existing  con- 
cepts and  existing  methods — has  been  my  enduring  but  never  sufficiently 
successful  resolution. 

Also  derived  from  medicine  were  consequential  convictions  respecting 

(1)  the  determining  importance  of  biochemical  occurrences — say  di- 
gestion,   assimilation,    metabolism,    excretion — after   the    organism   has 
finished  eating  and  the  interest  of  the  average  psychologist  has  faded; 

(2)  the  ultimate  scientific  value  of  systematic,  thorough,  and  detailed 
case  histories;  and  (3)  the  necessity  of  an  adequate  classification  of  the 
entities  and  processes  within  the  domain  of  one's  elected  discipline. 

I  have  nothing  more  to  say  under  the  present  heading  except  to 
avow  that  my  special  interest  in  the  dispositions  and  thoughts  (rather 
than  the  bodies)  of  human  beings  was  one  determinant  of  the  rather 
sudden  decision  I  made  to  shift  from  physiology  to  psychology.  Also 
influential  in  some  degree  were  the  impressions  ( 1 )  that  human  person- 
ality, because  of  its  present  sorry  state,  had  become  the  problem  of  our 
time — a  hive  of  conflicts,  lonely,  half-hollow,  half-faithless,  half-lost,  half- 
neurotic,  half-delinquent,  not  equal  to  the  problems  that  confronted  it, 
not  very  far  from  proving  itself  an  evolutionary  failure;  (2)  that  psycho- 

12  HENRY   A.    MURRAY 

analysis  had  already  made  appreciable  progress  in  exposing  and  interpret- 
ing the  deeper  processions  of  the  mind;  and  (3)  that  my  temperament 
was  more  suited  to  the  making  of  coarse  maps  of  newly  explored  areas 
than  to  the  refinement  of  relatively  precise  maps  of  familiar  ground. 


It  seems  that  I  was  scarcely  four  years  old  when,  like  a  cornerstone, 
the  law  was  laid  in  me  that  storytellers  should  begin  at  the  beginning. 
The  beginning  was  not  only  engaging  in  itself,  but  necessary  to  an 
appreciation  of  the  rest — all  succeeding  adventures,  stratagems,  con- 
flicts, loves,  and  triumphs  of  the  hero.  I  felt  with  Aristotle :  no  beginning, 
no  excitement  at  the  climax,  no  catharsis.  And  so,  if  my  father  or  my 
mother  failed  to  start  a  fairy  tale  with  "In  the  beginning/'  or  its 
equivalent,  "Once  upon  a  time,"  I  knew  that  I  was  about  to  be  deprived 
of  essential  information  and  this,  in  my  book  of  rules  and  regulations, 
was  ground  for  protest. 

But  more  consequential  than  this  early  requirement  for  a  good  fairy 
tale  was  my  first  down-to-earth  attempt  to  latch  on  to  the  beginning  of 
a  course  of  actual  events.  The  attempt  was  prompted  by  a  sudden  bellow 
that  originated,  I  soon  discovered,  from  a  strange  baby  in  my  parents' 
room.  Puzzled,  I  was  told  that  this  noisy  creature  was  my  brother  and 
perfectly  adorable.  Here  surely  was  a  notable  beginner;  but  what  was 
the  beginning  of  this  beginner?  My  inquiry  ended  with  the  answer  that 
Dr.  Anderton,  my  mother's  red-bearded  physician,  had  brought  him  in 
his  bag,  the  very  bag  from  which  I  had  so  often  seen  him  lift  spatula, 
swabs,  and  stethoscope. 

That  I  should  have  rested — I  won't  say  comfortably — with  the  words 
"doctor's  bag,"  that  I  should  have  abandoned  my  quest  for  basic  knowl- 
edge after  one  essay,  not  followed  the  path  of  my  intent,  the  path  of 
infinite  regressions,  one  leg  further  at  the  least,  a  step  which  would  have 
taken  me  to  the  place  where  Dr.  Anderton  obtained  the  babe,  that  I 
should  have  quit  so  soon,  is  evidence  of  a  docility  or  squeamishness  so 
unsuited  to  the  career  of  science  that  even  now  I  blush  to  acknowledge 
it  in  print.  If  all  along  I  have  been  stopped  at  the  very  verge  of  the 
unknown  by  some  constitutional  timidity,  it  is  possible,  yes  probable,  that 
I  have  failed  to  see,  or  to  interpret  properly,  or  to  report  candidly 
occurrences  that  were  beyond  the  stretch  of  well-established  scientific 
theories  or  beyond  the  bounds  of  embedded  moral  sentiments. 

I  have  mentioned  my  halt  at  the  "doctor's  bag"  conception  of  the 
fount  of  life — suggesting  parenthetically  that  I  might  not  have  Iain  down 
too  happily  with  this  solution — I  have  mentioned  this  defeat  of  curiosity 
as  a  possible  indirect  determinant  of  what  eventually  became  a  con- 

Preparations  for  Scaffold  of  a  Comprehensive  System      13 

firmed  interest  in  the  earliest  and  innermost  origins  of  things.  This  hypoth- 
esis might  help  to  explain  why,  twenty-five  years  later,  I  elected  to 
spend  the  greater  part  of  three  years  in  an  incubator  with  several  dozen 
eggs,  observing  and  measuring  the  chick  embryo's  earliest  manifestations 
of  vitality.  The  point  is  that  I  managed  at  long  last  to  get  inside  the 
doctor's  bag,  or,  better  still,  at  103.6°F,  almost  within  the  womb  of  the 
beginning  of  a  beginner.  Peering  through  a  microscope,  through  a  little 
fabricated  window  in  the  egg's  shell,  spellbound  as  any  libidinous 
voyeur,  I  witnessed  the  procession  of  momentous  transformations  that 
mark  the  hours  when  the  embryo  is  no  bigger  than  an  angel  perching 
on  a  pin  point.  Here,  it  seemed,  were  occurrences  of  great  significance 
which  into  concepts  no  contemporary  intelligence  could  digest. 

The  same  hypothesis  might  serve,  in  some  measure,  to  account  for 
my  disappointment,  if  not  aversion,  when  I  encountered  the  science  of 
psychology  at  college  and  listened  for  a  while  to  what  was  considered 
worth  announcing  about  the  perceptual  processes  of  the  adult  mind,  the 
mind  of  a  Western  intellectual,  a  mind  without  a  history,  strapped  to  a 
piece  of  apparatus  in  the  laboratory.  Also  in  keeping  with  this  hypoth- 
esis was  my  subsequent  embrace  of  Freud  with  all  his  facts  and 
legends  respecting  the  earliest  months  and  years  of  life.  Freud  kept  my 
first  commandment :  he  began  at  the  beginning.  In  my  initial  enthusiasm 
I  hardly  noticed  that  he  never  reached  the  consummation  of  the  allegory, 
the  heroic  adult  and  his  tragic  end. 

Depth  psychology  was  obviously  my  meat.  In  thedepths  one  came 
upon  the  earliest  and  most  determining  dispositions.  Whatever  initial 
doubts  I  had  respecting  unconscious  psychic  processes  were  soon  enough 
dispelled.  Several  weeks  with  Dr.  Jung  at  different  times,  three  years 
with  Dr.  Morton  Prince,  an  orthodox  Freudian  psychoanalysis,  and  a 
period  of  training  with  Dr.  Franz  Alexander  and  Dr.  Hanns  Sachs,  ten 
years  of  therapeutic  practice — these  experiences  were  hugely  influential  in 
shaping  my  personality  and  my  thought.  But  at  no  time,  to  the  annoyance 
of  my  friends,  was  I  a  good  Jungian,  a  good  Freudian,  a  good  Adlerian, 
or  a  good  schoolman  of  any  breed.  I  held  all  my  teachers  in  high  esteem, 
but  judged  that  each  of  them — necessarily  at  this  stage  of  theoretical 
development — was  more  or  less  one-sided.  The  notion  which  invited  me 
was  that  of  attempting,  with  the  aid  of  additional  ideas  culled  from  the 
writings  of  McDougall,  of  Lewin,  and  of  my  colleagues  at  Harvard,  a 
preliminary  revision  and  integration  of  current  academic  and  psycho- 
analytic theories  to  accord  with  a  large  collection  of  reasonably  solid 
facts  obtained  by  the  multiform  method  of  assessment.  This  effort  re- 
suited  in  the  crude  blueprint  for  a  system  which  a  number  of  us  sub- 
mitted in  Explorations  in  Personality  [3],  a  blueprint  which  stressed  the 
earliest  and  least  accessible  determinants  of  behavior.  We  did  not  do 
this  to  the  satisfaction  of  the  psychoanalysts,  first,  because  all  behavior 

14  HENRY   A.    MURRAY 

was  not  traced  to  infantile  sexuality  and  aggression  and  second,  because 
we  classified  overt  actions  as  they  occurred,  whether  or  not  we  had 
reasons  to  believe  that  they  were  subsidiary  to  deeper,  hidden  aims. 

But  now,  if  I  may,  I  shall  mention  another  disgrace  of  childhood 
which  seems  relevant  to  this  topic — what  I  have  called  my  interest  in 
origins  and  'beginnings.  If,  awhile  back,  I  almost  disqualified  my  child- 
hood self  as  a  potential  truth  seeker,  by  mentioning  that  moment  of 
scarcely  pardonable  poltroonery  in  the  face  of  the  Great  Riddle,  what 
I  now  have  to  confess  is  evidence  of  something  bordering  on  complete 
damnation  in  the  scales  of  science.  Not  going  beyond  Dr.  Anderton  and 
his  bag  signalized  a  defect  in  daring  and  determination  to  solve  prob- 
lems; but  worse  than  this  is  the  inability  to  know  a  pithy  problem  when 
you  see  it.  So  far  as  I  can  recall,  if  truth  will  out,  I  was  never  prompted 
to  ask  about  the  very  beginning,  the  beginning  of  mankind  or  the 
beginning  of  the  world.  Passively  and  without  suspicion  or  comment  I 
received  the  news  that  some  six  thousand  years  ago  God — who,  in 
pictures  I  was  shown,  had  a  big  beard,  not  red  like  Dr.  Anderton's,  but 
white  as  my  venerable  and  remote  grandfather's — that  God  had  con- 
structed the  first  man  in  a  single  day,  and,  a  little  later,  molded  from  one 
of  this  man's  ribs  the  first  woman,  et  cetera,  et  cetera. 

I  suspect  that  it  was  the  swallowing  and  digesting  of  this  fable,  trust- 
1  fully  and  without  complaint,  which  determined,  to  some  degree,  my 
gust  for  Darwin  and  the  evolutionists  who  succeeded  him,  as  well  as  the 
joy  I  felt  in  shedding  the  constraining  creeds  of  orthodox  religion.  It 
was  as  though  a  strait  jacket  had  been  unfastened  and  I  stepped  out  to 
breathe  and  move  and  think  for  the  first  time  without  embarrassment. 

It  was  from  biology  and  chemistry  that  I  received  the  exciting  notion 
that  man  is  descended  from  the  very  humblest  of  parents,  a  more  or 
less  fortuitous  combination  of  chemical  elements — such  low-caste  stuff 
as  hydrogen,  oxygen,  carbon,  and  nitrogen — and  that,  instead  of  a  day, 
it  took  two  billion  years  or  more  to  shape  him.  Also  noteworthy  was  the 
evidence  that  the  wondrous  evolutions  of  man  and  of  his  productions 
may  be  credited,  in  some  measure,  to  the  very  tendency  which  in  the 
Garden  of  Eden  version  led  to  his  disgrace  and  fall,  that  is,  the  inborn 
tendency  to  explore  and  to  experiment  among  forbidden  things. 

My  enthusiasm  for  this  theory  becomes  more  intelligible  when 
viewed  in  conjunction  with  the  next  orienting  disposition  to  be  listed. 


It  is  hard  to  decide:  should  I  speak  here  of  a  predisposition  that 
sensitized  me  to  a  certain  class  of  facts  or  should  I  speak  of  a  certain 

Preparations  for  Scaffold  of  a  Comprehensive  System      15 

class  of  facts  which  engendered  a  disposition  to  accept  them  and  look 
for  more  of  the  same  kind?  I  have  always  thought  it  good  emotional 
policy  not  only  to  enjoy,  so  far  as  possible,  the  inevitable,  but  to  will 
the  obligatory.  In  this  case,  however,  I  am  inclined  to  stress  the  inner 
bent  ahead  of  the  compelling  facts  because  to  the  majority  of  psycho- 
logical theorists  these  facts  are  not  particularly  compelling. 

I  am  referring  to  facts  which  particulary  attracted  me  during  my 
studies  at  the  Rockefeller  Institute  of  the  physiological  ontogeny  of 
chicken  embryos.  To  summarize  a  long  story,  what  seemed  both  most 
obvious  and  most  important  about  the  interior  of  the  embryo  were  (a) 
the  givenness,  the  inherent  spontaneity,  of  its  cellular  activities  and  (b) 
the  continuous  sequence  of  orderly  metamorphoses  (clearly  perceptible 
under  the  microscope)  which  resulted  from  these  activities,  and  hence 
the  necessity  of  including  formative  (constructive)  processes  in  one's 
scheme  of  variables. 

Unintrusive  observation  was  enough  to  nail  down  the  self-evident 
proposition  that  chemical  and  physical  activity,  metabolism  and  move- 
ment, are  integral  properties  of  every  animate  body,  things  to  be  in- 
cluded in  the  very  definition  of  life.  Also,  it  appeared  that  organic 
processes  are  not  only  primarily  endogenous,  autonomous,  and  proactive 
(initiated  and  sustained  from  within,  rather  than  being  merely  reactive 
to  external  stimuli)  but  especially  in  the  early  stages  of  development 
are,  so  far  as  one  can  see,  not  perfectly  coordinated  with  other  processes, 
not  constantly  directed  toward  the  achievement  of  effects  extrinsic  to 
themselves.  This  view  of  things  was  initially  implanted  by  observing, 
time  after  time,  the  very  first  beat  of  the  uncompleted  embryonic  heart 
and  noting  that  it  contracted  irregularly  and  then  regularly  for  quite 
a  while — I  forget  how  long  precisely — before  the  blood  vessels  and  the 
corpuscles  were  far  enough  along  in  their  development  to  make  it 
possible  for  this  organ  to  perform  its  predestined  function,  namely,  to 
pump  oxygen-refreshed  blood  through  the  arteries  of  the  body.  The 
primitive  heart  was  merely  exhibiting  its  emergent  capacity  to  contract, 
like  a  playful  child  or  puppy,  achieving  for  a  period  no  effects  outside 
its  own  growth  of  form  and  potency. 

This  notion  of  endogenous,  initially  undirected  and  uncoordinated, 
process-activity  constrained  me,  in  later  years,  on  the  one  hand,  to 
qualify  my  acceptance  of  the  fashionable  stimulus-response  formula, 
with  its  implicit  assumption  of  a  nothing-but-reactive  organism  and,  on 
the  other  hand,  to  qualify  my  acceptance  of  the  proposition  that  all 
activities  are  motivated.  In  short,  I  take  "life" — say,  the  ceaseless 
processes  of  metabolism — as  given,  just  as  Newton  took  motion  as  given, 
and  do  not  look  for  something  antecedent  to  it,  except  in  an  evolutionary 

16  HENRY  A.    MURRAY 

The  other  influential  impression  I  received  from  my  studies  of 
embryonic  physiology  was  that,  during  the  first  phases  of  its  career,  a 
relatively  large  proportion  of  the  totality  of  processes  within  a  living 
organism  is  involved  in  the  development  of  somatic  substance,  in  the 
work,  let  us  say,  of  anabolism,  of  incorporating  and  combining  new 
elements,  and  so  of  constructing  and  of  reconstructing  parts  unexampled 
in  the  history  of  that  particular  unity  of  animation.  In  other  words,  the 
most  significant  characteristic  of  the  embryo  is  not  so  much  the  arrange- 
ment of  its  perceptible  component  forms  of  matter  at  this  or  that 
moment,  as  its  activity  in  forming  and  transforming  forms  of  matter. 
Defining  "energy53  as  the  capacity  to  produce  change,  change  of  relations, 
we  can  say  that  most  of  the  energy  of  the  embryo  is  devoted  to  generative 
changes,  that  a  host  of  processes  precede  forms,  one  of  the  effects,  or 
"functions,"  of  some  processes  being  to  build  and  to  rebuild  them.  That 
is,  the  organism,  being  an  open  system  (as  Bertalanffy  showed  me  later), 
selects  from  its  environment,  incorporates  and  synthesizes,  potentially 
energic  matter,  and  thereby  increases  its  resources,  taking  a  course 
opposite  to  that  defined  by  the  second  law  of  thermodynamics  (which 
applies  to  closed  systems).  Furthermore,  clearly  perceptible  under  the 
microscope  were  divisions  of  the  soma  into  regions  and  in  each  region 
the  production  of  distinctive  structures,  in  short,  morphological  dif- 
ferentiation, preparatory  to  specialization  of  functions. 

It  was  these  observations  of  embryonic  developments,  besides  what  I 
could  understand  about  the  science  of  energetics,  which  initially  pre- 
disposed me  to  stress  "mythologies53  of  energy,  process,  change,  function, 
more  than  "mythologies55  of  matter,  structure,  permanence,  and  to  regard 
the  organism  as  ordered  successions  of  different  kinds  of  processes,  the 
effects  of  some  of  them  being  primarily  internal — formations  and  re- 
formations of  component  structures — with  a  re-ordering  of  the  processes 
occurring  consequentially.  In  short,  according  to  this  way  of  thinking, 
creativity  is  an  inherent  property  of  the  organism  and  stability  is  another. 

Four  of  the  ultimate  resultants  of  my  interest  in  process,  develop- 
ment, and  creativity  were  ( 1 )  the  adoption  of  the  whole  history  of  an 
organism,  the  entire  life  span  of  a  personality,  as  the  macro-temporal 
unit  that  requires  formulation  (although  it  may  be  half  a  century  before 
a  satisfactory  way  of  doing  this — an  adequate  conceptual  scheme  and 
an  adequate  methodology — is  devised) ;  (2)  an  interest  in  all  manifesta- 
tions of  significant  changes  of  personalities — progressive  transformations, 
eliminations,  and  reconstructions,  learnings,  extinctions,  and  relearnings, 
regressions  and  deteriorations — and  in  the  determinants  of  such  changes, 
and  hence  a  special,  but  by  no  means  exclusive,  emphasis  upon  the  in- 
fluential experiences  of  childhood;  (3)  a  devotion  to  all  forms  of  the 
imagination — dreams,  fantasies,  prospections,  ordinations  (plans),  plays, 

Preparations  for  Scaffold  of  a  Comprehensive  System      17 

story-constructions,  myths,  rituals,  religious  conceptions,  works  of  art, 
and  scientific  speculations — as  manifestations  of  involuntary  and  largely 
unconscious  process-activities  which,  when  influenced  by  a  strong  and 
continuing  intention,  may,  in  some  cases,  have  a  definitely  creative  out- 
come; and  (4)  the  construction  of  a  large  number  of  methods  (most 
of  them  unpublished)  for  the  eduction  and  exposure  of  imaginal  proc- 
esses and  products  (so-called  projective  tests) . 

Imaginal  processes  and  products  appeal  to  me  not  only  because  of 
their  intrinsic  interest,  but  because  they  have  been  shown  to  be  the  best 
source  of  dependable  clues  of  underlying  (and  often  unconscious)  dis- 
positions and  conflicts  of  dispositions.  An  often  verified  hypothesis  is 
that  some  of  these  inferable  dispositions  are  residua  of  deformative  in- 
fantile experiences  and  that  a  few  of  them  are  prodromes  of  conditions 
in  the  offing. 



It  was  in  1920,  during  my  studies  of  oscillations  of  the  physicochem- 
ical  equilibria  in  the  blood,  under  the  tutelage  of  Lawrence  J.  Hender- 
son, author  of  The  Order  of  Nature  [1],  that  I  first  became  familiar 
with  the  organismic,  or  organismal,  proposition,  as  formulated  by  E.  S. 
Russell  in  1916  and  elaborated  by  W.  E.  Ritter  in  1919.  Belief  in  its 
essential  validity  was  confirmed  a  little  later  (sometime  before  I  heard 
anything  about  gestalt  psychology)  by  observations  of  the  embryo — per- 
ceiving the  sequential  effects  of  Spemann's  genetical  "organizers" — and 
by  studies  of  the  regulatory  functions  of  the  autonomic  nervous  system 
in  conjunction  with  the  endocrines.  Clearly  demonstrable  in  higher 
animals  are  vertical  integrations  of  superordinate  and  subordinate  loci 
of  control,  levels  of  directors  and  coordinators,  "lines  of  command" 
starting  from  some  center  in  the  segmented  neuraxis,  or  lower  brain, 
and  ending  in  regional  plexuses  and  local  nervous  networks,  a  hier- 
archical system,  depending  on  "feedbacks"  (as  we  say  today),  which 
executes  the  genetically  determined  "domestic  policy"  of  the  organism. 

Here  it  might  be  appropriate  to  refer  to  Cannon's  principle  of 
homeostasis,  and  to  the  fact  that  consideration  of  the  radical  develop- 
ments during  the  embryonic  period  led  me  to  stress  the  concept  of 
progressive  disequilibrium,  continuity  through  expansive,  constructive 
change,  as  a  supplement  to  that  of  homeostasis  (which  is  more  applicable 
to  the  interior  activities  of  adult  organisms).  The  concept  of  homeostasis 
(the  maintenance  and,  if  disturbed,  the  restoration  of  the  same  state 

18  HENRY   A.    MURRAY 

of  equilibrium)  is  a  basic  scientific  induction,  defining  as  it  does  the 
measurable  relationships  of  multifarious  interdependent  elements  and 
processes,  relationships  which  either  persist  unchanged,  or,  if  modified 
by  some  intruding  element  of  exigency,  are  in  due  course  reestablished. 
But  it  should  be  noted  that  this  principle,  as  commonly  defined,  is  valid 
only  within  a  certain  time  span.  The  time  span  varies  with  the  age  of 
the  organism  as  well  as  with  the  system  (physiological,  psychological,  or 
sociological)  that  is  under  consideration.  In  the  body  of  a  healthy  adult, 
the  morphological,  physicochemical,  and  physiological  relationships  are 
quite  stable,  or  soon  restabilized,  over  a  period  of  many  years  despite 
the  slow  changes  which  eventually  result  in  the  signs  and  symptoms  of 
senescence.  But  in  the  embryo  homeostasis  has  virtually  no  span  at 
all,  or  an  extremely  short  one:  the  organism  as  a  system  being  char- 
acterized in  all  its  manifestations  by  perpetually  changing  states  of 
equilibria,  states  that  move  in  an  irreversible  direction.  In  short,  the 
embryo  is  in  disequilibrium  or,  at  most,  transitional  equilibrium  from  first 
to  last.  Comparable,  I  thought  later,  though  less  striking  to  the  eye  and 
less  susceptible  to  precise  measurement,  are  the  seasons  of  transitional 
equilibrium  on  the  psychological  level,  which  occur  most  obviously  in 
childhood  but  also  later,  during  the  early  phases  of  some  new  enterprise, 
let  us  say,  or  when  the  creative  imagination  is  steadily  advancing.  At 
such  times  psychological  processes  are  transformative,  and  when  they 
terminate,  the  person  is  a  different  person,  or  his  sphere  of  relationships 
is  different,  and  there  is  a  different  equilibrium  to  be  sustained. 

Although  I  came  away  from  my  embryological  studies  with  a  firm 
belief  in  the  unity  of  the  organism  through  change,  in  orderly  dif- 
ferentiations and  integrations,  my  medical  training  had  established  a 
special  vigilance  in  respect  to  signs  and  symptoms  of  functional  imper- 
fection, and  I  soon  discovered  how  normally  abundant  are  such  evidences 
on  the  psychological  level,  evidences  of  disunity,  of  retardation,  deviation, 
deformation,  and  retrogression.  It  appears  that  millions  of  years  of 
evolution  have  resulted,  on  the  one  hand,  in  an  almost  perfect  system, 
let  us  say,  of  somatic  operations,  and  on  the  other  hand,  in  a  human 
brain  which  contains  at  birth  no  comparably  ordered  system  of  depend- 
able proactions  and  reactions,  but  instead,  a  matrix  of  potentialities  in 
a  relatively  amorphous  state,  potentialities  for  unprecedented  develop- 
ments of  talent,  at  one  extreme,  and  for  idiocy  and  lunacy,  at  the  other. 
Hence,  especially  for  human  beings,  life  is  a  continuous  procession  of 
explorations,  surmises,  hunches,  guesses,  and  experiments,  failures  and 
successes,  of  learnings  and  relearnings — aging  consisting  of  a  sequence 
of  gradual  and  occasionally  abrupt  indurations  (rigidifications,  solidifica- 
tions,  fixations,  hardenings),  both  of  forms  and  functions.  Consequently, 
a  psychologist  has  to  deal  conceptually  with  doubt,  distrust,  indecision. 

Preparations  for  Scaffold  of  a  Comprehensive  System      19 

and  postponement  of  behavior  among  his  subjects,  together  with  oc- 
currences and  continuities  of  competition  and  paralyzing  conflict  be- 
tween their  dispositions. 

My  bent  toward  organismic,  holistic,  molar,  or  "gestalt"  con- 
ceptions of  the  personality  and  its  activities  scarcely  fitted  me  to  wax 
avid  when  I  encountered,  later  on,  the  then  dominant  elementalistic, 
connectionistic,  chained-reflex,  molecular  theories  of  learning,  theories 
that  were  being  hungrily  ingested  by  all  who  cared  on  what  side  their 
academic  bread  was  buttered.  Of  the  two  fallacies,  reductive  and  seduc- 
tive, so  nicely  discriminated  by  Herbert  Feigl,  I  was  more  liable  to  the 
second,  though  I  had  no  use  for  those  lazy  white  elephants  of  the  mind — 
huge,  catchall,  global  concepts  signifying  nothing.  Eventually  I  was 
persuaded  by  Professor  Boring — more  generous  of  his  time  than  any 
teacher  I  ever  had — that  the  principles  of  elementalism  and  associa- 
tionism  are  applicable  under  many  circumstances,  especially,  let  us  say, 
to  the  establishment  of  certain  neurotic  symptoms  as  well  as  to  condi- 
tionings that  occur  below  the  level  of  conscious  control  or  when  the 
mind  is  tired  or  confused  and  functioning  below  par.  In  short,  elemen- 
talism (emphasis  on  parts,  integrants,  components)  and  holism  (em- 
phasis on  wholes,  integrates,  ordinations  of  components)  are  necessary 


One  passes  by  inseparable  gradations  from  an  interest  in  the  auto- 
nomic-endocrinal  coordination  of  the  multifarious  somatic  processes  of 
the  body  and  in  the  local  effects  of  their  different  operations,  to  an 
interest  in  the  cortical  ordination  of  sensory,  muscular,  and  verbal 
processes  toward  successive  achievements  of  different  overt  effects,  most 
of  which  endeavors,  if  successful,  contributing  in  some  way  or  other  to 
the  well-being  of  the  total  organism. 

Hence,  it  was  already  in  the  cards  I  held  that,  on  entering  the  do- 
main of  psychology,  I  should  very  soon  become  concerned,  not  so  much 
with  reflexes  and  patterns  of  muscular  movements,  as  with  the  various 
changes  effected  by  such  movements  and  the  changes  in  the  states  and 
thoughts  of  other  people  effected  by  spoken  words  and  sentences. 

The  fundamental  fact,  it  seemed  to  me,  is  the  survival  of  the  living 
organism,  the  continuation  of  its  metabolic  processes,  and  the  dependence 
of  this  procession  upon  the  periodic  attainment  of  a  number  of  distinct 
effects,  such  as  the  inspiration  of  oxygen,  the  expiration  of  carbon 
dioxide,  the  ingestion  of  water  and  food,  and  the  excretion  of  waste 
products.  The  different  processes,  modes,  and  subeffects  whereby  the 

20  HENRY   A.    MURRAY 

same  kind  of  terminal  effect  is  achievable  in  different  species  of  organisms 
or  even  in  the  same  organism  at  different  times  were,  at  the  start,  a 
matter  of  considerable,  but  subsidiary,  importance.  A  man  shivering 
with  cold  may  restore  optimal  body  temperature  by  moving  to  a  warm 
place,  putting  on  an  overcoat,  closing  doors  and  windows,  lighting  a 
fire,  turning  on  the  radiator,  taking  a  hot  drink,  or  exercising  strenuously. 
Defined  in  terms  of  physical  vectors  (locomotions  or  manipulations  in 
space)  these  are  different  actions,  but  the  beneficent  effect  in  all  cases  is 
the  same.  Indeed,  a  person  may  obtain  all  necessary  "goods"  with  the 
minimum  of  activity  on  his  part:  they  may  be  furnished  providentially 
by  nature,  gratuitously  by  parents  or  friends,  or  in  exchange  for  money, 
by  domestic  servants  or  employees.  I  had  a  good  friend  who  lay  in 
bed,  blind  and  completely  immobilized  from  his  neck  down,  for  twenty 
years.  He  had  a  sensitive  and  brilliant  mind  that  was  bubbling  over 
with  unimpaired  effectiveness  and  charm  until  his  death,  and  yet  he  saw 
nothing  and  never  moved  a  muscle.  Every  act  necessary  to  his  survival, 
to  the  stimulation  of  his  feelings,  and  to  the  increase  of  his  knowledge 
had  to  be  performed  by  someone  else.  This  was  but  one  of  countless  ob- 
servations which  persuaded  me  of  the  necessity  of  providing  concepts  for 
the  analytical  dissection,  whenever  necessary,  of  any  short  segment  of 
activity  into  ( 1 )  kinds  of  exciting  initial  situations,  ( 2 )  kinds  of  processes 
(e.g.,  covert  psychic  processes,  overt  psychomotor  or  psychoverbal 
processes)  with  or  without  kinds  of  utilities  (e.g.,  tools,  weapons,  con- 
veyances, telephone,  typewriter,  etc.),  (3)  kinds  of  modes,  or  styles,  of 
processional  activity,  kinds  of  psycho-expressive  processes  (e.g.,  speed, 
grace,  gestures  or  tone  of  voice  expressive  of  uncertainty,  anxiety,  self- 
confidence,  anger,  good  will,  deference,  compassion,  etc.),  and  (4)  kind 
of  effect  (change  from  the  initial  exciting  to  the  terminal  gratifying 

In  my  persistent  efforts  to  move,  step  by  step,  toward  an  adequate 
solution  to  such  problems,  I  was  greatly  assisted  by  the  reported  observa- 
tions and  formulations  (1)  of  biologists  from  Darwin  on,  and  of 
others,  particularly  McDougall,  who  had  used  the  concept  of  instinct 
as  their  tool;  (2)  of  Freud  relative  to  the  sex  instinct,  aggression,  and 
anxiety,  and  of  Adler  relative  to  the  craving  for  superiority;  (3)  of 
Tolman  and  other  animal  psychologists  who  had  carried  forward  the 
endeavor  to  define  and  measure  rigorously  different  drives;  (4)  of  Lewin 
with  his  constructs  of  tension  system  and  of  quasi  need;  and  (5)  of 
sociologists  regarding  the  wants  of  men  for  status  and  for  power. 

In  Explorations  of  Personality  I  attempted  to  define  a  number  of 
actional  dispositions  which,  in  the  absence  of  a  less  objectionable  desig- 
nation, were  termed  "needs"  (or  "drives").  These  constructs  proved  use- 
ful in  categorizing  inferentially  the  overt  behaviors  of  the  subjects  we 

Preparations  for  Scaffold  of  a  Comprehensive  System      21 

studied  as  well  as  the  behaviors  of  the  characters  in  the  stories  they  com- 
posed. But  this  particular  working  inventory  of  human  drives  (kinds  of 
motivations,  purposes,  intended  effects,  goals)  was,  in  several  respects, 
deficient,  and  ever  since,  these  deficiencies  have  kept  provoking  me  to 
prolonged  efforts  to  conceive  of  fitting  remedies.  An  account  of  today's 
resultant  of  my  arduous  and  still  continuing  endeavor  to  arrive  at  a 
more  comprehensive  and  integrated  system  will  be  presented  in  a  later 

Before  leaving  this  topic  I  should  say  that  I  have  not  been  satisfied 
to  limit  my  objective  to  the  formulation  of  overt  behaviors,  certainly  not 
to  the  formulation  of  purely  physical  behaviors.  Indeed,  after  perceiving 
that  the  food-ingesting  activities  of  animals  and  of  men  are  not  at  all 
representative  of  the  majority  of  human  actions  (as  Maslow  has  pointed 
out)  but,  being  most  readily  formulated  in  physical  terms,  are  repeatedly 
used  nonetheless  to  illustrate  this  and  that  concept  or  generalization  or 
to  serve  as  foundation  for  this  and  that  postulational  system,  and  that 
they  thus  constitute  an  alluring  conceptual  trap  for  the  unwary  theorist 
— perceiving  all  this,  I  established  in  myself  a  prohibition  (which  I 
guiltily  break  occasionally)  against  using  the  hunger  drive  and  its  en- 
suing motor  patterns  and  effects  as  paradigm  of  directional  behaviors  or 
even  as  a  reliable  reference  point  for  speculation. 

As  I  see  it,  a  psychologist  should  be  concerned  not  only  with  the 
formulation  of  overt  interpersonal  verbal  communications,  the  imme- 
diate (intended)  effects  of  which  are  changes  of  some  kind  among  the 
dispositions,  evaluations,  represented  facts,  interpretations,  or  commit- 
ments of  the  other  person,  but  also  with  the  formulation  of  covert  intra- 
verted  mental  activities,  the  immediate  (intended)  effects  of  which  are 
such  things  as:  a  better  interpretation  and  explanation  of  some  recalled 
event  or  of  some  current  physical  symptom,  a  reevaluation  of  one's  own 
enactions  (past  behaviors)  or  present  abilities,  the  definition  of  the  con- 
tent and  boundaries  of  a  required  concept,  the  composition  of  the  plot 
of  a  story  to  be  written,  the  resolution  of  a  conflict  between  two  purposes, 
or  the  ordination  of  a  plan  of  action  (tactics)  to  be  executed  at  some 
future  date. 


I  owe  much  to  the  incomparable  Alfred  North  Whitehead  and  the 
incomparable  Kurt  Lewin,  nothing  less  than  the  conviction  that  con- 
crete reality  is  to  be  found  only  in  the  momentary.  With  theoretical 
physics  in  mind,  Lewin  devoted  a  good  deal  of  his  unusual  imaginative 
powers  to  the  definition  of  space  constructs,  topological  and  hodological, 



the  momentary  field;  whereas  Whitehead,  founding  his  penetrating  re- 
flections on  organic  and  mental  phenomena,  emphasized  the  momentary 
process^  the  perpetual  becoming  and  perishing  of  "actual  occasions"  and 
the  historic  continuity  or  progression  of  these  occasions.  Although  I  have 
never  gained  sufficient  understanding  of  Whitehead's  terminology  to 
apply  the  categorial  scheme  of  his  philosophy  of  organism  to  the  realm 
of  ordinary  human  experience  and  behavior,  I  am  indebted  to  him  for  a 
number  of  conceptions  which  I  have  revised  to  suit  the  purposes  of  a 
psychologist.  First  among  these  is  the  concept  of  an  event,  or  fact,  as  a 
participation  of  processes  in  which  two  or  more  interdependent  entities 
are  involved  occurring  in  a  certain  place  or  along  a  certain  path,  within 
a  certain  medium,  through  a  certain  segment  of  time,  and  resulting  in 
a  certain  kind  of  change.  I  conceive  of  a  range  of  events  of  different 
molarities.  Theoretically,  an  ultimate  submicro  event  would  have  the 
smallest  spatial  scope   (smallest  containing  field),  the  smallest  entity 
scope  (fewest  component  particles),  smallest  process  scope  (fewest  dis- 
tinguishable changes),  and  shortest  temporal  span  (duration).  (For  ex- 
ample, it  is  estimated  that  tau  and  theta  mesons  are  composed  and  de- 
composed in  about  a  hundred-millionth  of  a  second. )  Some  micro  events 
occur  within  the  boundaries  of  solids,  i.e.  entities  that  can  be  treated  as 
solids  under  most  conditions  (anything  from  a  crystal  to  a  planet),  but 
others  are  integrated,  synchronously  and  sequentially  in  time,  in  such  a 
way  as  to  constitute  an  event  of  greater  scope  and  span,  and  this  event, 
in  turn,  can  be  seen  to  constitute  a  necessary  part,  or  phase,  of  an  event 
of  still  greater  scope  and  span,  something  that  can  be  defined  in  terms 
of  a  single  resultant  process — secretion  of  a  hormone  by  one  cell,  a  single 
color  sensation,  influxion  of  a  single  image,  contraction  of  a  single  mus- 
cle fiber — or  in  terms  of  a  longer  or  more  massive  process — secretion  of 
saliva,  contraction  of  the  heart,  perception  of  a  configuration,  momen- 
tary feeling,  evaluation  of  an  object,  movement  of  a  limb,  etc.  Such  an 
event  may  be  a  part  of  a  yet  larger,  longer  whole — say,  a  stimulus- 
response  unit  (perception,  apperception,  and  evaluation  of  a  pertinent 
entity,  concurrent  emotion,  actuation  of  a  pattern  of  muscular  move- 
ments against  resistance,  production  of  an  effect,  perception,  appercep- 
tion, and  evaluation  of  this  effect) .  Thus,  by  increasing  step  by  step  one's 
scope  and  span  of  concern,  one  arrives  at  the  largest  and  longest  defin- 
able unit  of  activity,  a  macro  event.  A  personologist  usually  has  to  deal 
with  macro  events,  or  proceedings;  and  from  the  fullness  of  each  of  these 
he  abstracts  those  variables  which  are  relevant  to  his  purpose,  in  the 
knowledge  that  numberless  other  variables  will  be  unrecorded  and  hence 
omitted  from  his  formulation.  Thus,  the  major  concepts  of  the  scaffold 
to   be  built — such  as  need,   entity,   configuration,   process,  succession, 
effect,  place,  route,  time— are  all  considered  to  be  abstractions  from  an 
event  or  progression  of  events. 

Preparations  for  Scaffold  of  a  Comprehensive  System      23 

As  twentieth  century  inhabitants  of  the  Western  world,  we  seem  to 
be  living  and  acting — partly  as  a  consequence  of  our  acquired  Indo- 
European  language — in  euclidean  space,  moving  about  on  the  support- 
ing surface  of  an  assumptively  permanent  material  planet  amid  a  great 
variety  of  substantial  objects,  inanimate  and  animate,  natural  and  arti- 
ficial (man-made),  some  transient,  some  relatively  permanent,  each  with 
its  distinguishing  physical  attributes.  "Such  presumptions,"  as  Whitehead 
says,  "are  imperative  in  experience,"  and  "in  despite  of  criticism,"  we 
still  employ  them  "for  the  regulation  of  our  lives."  And  so,  for  better  or 
for  worse,  I  too  have  employed  them,  not  only  in  the  regulation  of  my 
life  but,  with  certain  qualifications,  in  the  regulation  of  my  theorizings. 

If  I  were  forced  to  choose  one  side  of  the  age-old  antinomy  between 
the  "metaphysics  of  substance"  and  the  "metaphysics  of  flux,"  my  tem- 
perament would  decide  in  favor  of  the  latter,  the  version  of  the  universe 
that  is  linked  in  our  minds  with  some  vivid  sentences  attributed  to  Hera- 
clitus.  But,  as  I  see  things  and  events,  it  is  not  necessary  to  go  to  one 
side  or  the  other,  either  of  this  classical  division  between  different  aspects 
of  nature  or  of  other  dichotomies,  such  as  that  between  matter  in  space 
and  motion  in  time,  or  between  instantaneous  configurations  of  material 
bodies  and  modifications  of  these  configurations,  or  between  chemical 
structure  and  chemical  properties  and  processes,  or  between  form  and 
function,  or  between  anatomy  and  physiology,  or  between  entity  and 
activity,  or  between  actor  and  action,  or  between  noun  and  verb.  It  is 
possible  to  choose  both  sides  and  combine  them  in  single  propositions. 

Perhaps  ray  most  influential  basic  model  is  that  of  biochemical 
metabolism,  repetitive  and  restorative  as  well  as  progressively  and  irre- 
versibly transformative:  the  lifelong  succession  of  compositions,  decom- 
positions, and  re  compositions  of  concrescences  and  perishings,  of  vital 
chemical  substances.  Here  is  incessant  flux  certainly,  with  the  catabolism 
of  anabolized  materials  liberating  the  energy  for  every  manifestation  of 
vitality  (thermal,  chemical,  electrical,  mechanical — emotional,  disposi- 
tional,  mental,  and  muscular) ;  and  here  also  are  countless  instantaneous 
configurations  of  substances  within  cells,  of  cells  within  organs,  and  of 
organs  within  a  body,  some  parts  of  which  (skeleton,  ligaments,  con- 
nective tissue,  skin)  are  relatively  solid  and  enduring  like  the  framework 
of  a  house.  Consideration  of  anabolisms,  in  which  two  or  more  chemical 
entities  combine  to  form  or  to  re-form  a  more  complex  entity,  where 
one  can  attribute  the  course  of  events  to  no  single  actor  and  his  act,  has 
led  me  to  conceptualize,  in  many  cases,  systems  of  participant  entities 
and  participating  processes  rather  than  placing  the  major  burden  of  de- 
termination on  one  person  or  on  one  person's  conscious  purpose.  Here 
one  might  think  of  the  mental  participations  involved  in  creative  activity, 
with  conscious  intention  playing  but  a  minor  role,  or  of  the  emotional, 
verbal,  and  actional  participations  of  two  lovers. 

24  HENRY  A.    MURRAY 

At  this  point  let  me  explain  for  clarity's  sake  that  in  view  of  the 
mind's  tendency  to  "spatialize"  everything,  as  Bergson  pointed  out,  and 
in  view  of  the  ambiguous  usages  in  the  social  sciences  of  such  words  as 
structure,  configuration,  form,  pattern,  integration,  etc.,  I  prefer  to  re- 
strict the  word  configuration  to  the  instantaneous  (transient),  and  the 
word  structure  to  the  enduring,  spatial  relations  of  the  substantial  com- 
ponents of  an  entity,  assemblage  of  entities,  or  region  (extended  surface 
area),  and  to  use  the  word  succession  to  designate  the  once-occurring, 
and  the  word  integration  to  designate  the  recurrent  temporal  relations 
of  the  component  processes  of  a  proceeding  (uninterrupted  activity, 
endeavor,  interaction).  According  to  this  terminology  it  would  be 
proper  to  speak  of  the  structure  of  a  house,  of  a  painting,  of  an  organ- 
ism, of  a  chemical  compound,  of  a  crystal,  or  of  an  atom;  and  it  would 
be  proper  to  speak  of  the  integration  of  mechanical,  electrical,  chemical, 
mental,  verbal,  or  musical  processes,  through  a  certain  period  of  time. 
One  could  also  speak,  in  a  highly  abstract  way,  of  the  hypothetical 
structure  of  the  mind  or  of  the  personality,  although  mind  and  person- 
ality are  known  to  us  only  through  successions  of  covert  (subjective) 
and  overt  (objective)  processes.  You  see  I  am  wary  of  the  word  "struc- 
ture," because,  if  used  to  describe  concatenations  of  activities  one  gets 
that  impression  of  permanence,  regularity,  and  lawfulness  which  is  so 
dear  to  the  hearts  of  scientists  and  yet  so  incongruent  with  the  facts  in 
many  instances. 

The  debt  I  owe  to  Lewin  can  be  most  simply  set  forth  if  we  restrict 
thought,  for  the  time  being,  to  the  motor  activities  of  one  person  from 
the  beginning  to  the  end  of  a  single  simple  proceeding,  or  endeavor,  a 
goal-directed  and  goal-attaining  course  of  action.  In  such  a  case  the 
"whole"  effect  (attained  goal)  of  the  pattern  of  muscular  processes  can 
be  defined  by  designating  the  relevant  differences  between  the  structure  of 
the  physical  field  at  the  initiation  of  the  activity  and  the  structure  of  the 
physical  field  at  its  termination.  This  will  tell  us  what  the  person  "did" — 
he  moved,  let  us  say  roughly,  from  one  location  to  another,  or  moved 
an  object  from  a  table  to  his  mouth,  or  put  a  new  tire  on  his  car,  or  hung 
a  picture  over  his  desk,  etc.  But  more  than  this,  ideally  considered — 
and  here  is  where  Lewin  comes  in — a  sufficient  characterization  of  the 
field  at  the  start  of  the  activity,  and  at  every  instant  from  then  on,  would 
set  forth  the  immediate  determinants  of  the  over-all  direction  of  the 
activity  as  well  as  of  each  successive  part,  or  unit,  of  the  whole.  As 
Lewin  put  it,  "the  behavior  b  at  the  time  t  is  a  function  of  the  situation 
S  at  the  time  t  only,"  where  S  denotes  the  total  situation  (field) — the 
field  of  forces  within  the  person  (internal  situation)  as  well  as  the  field 
of  forces  exterior  to  the  person  (external  situation),  as  apperceived  and 
evaluated  by  the  S,  The  initiating  total  field  (a  momentary  cross  section, 

Preparations  for  Scaffold  of  a  Comprehensive  System      25 

or  time-slice,  through  everything  that  is  influential)  determines  the  be- 
havior resulting  in  the  next  field,  which,  in  turn,  determines  the  behavior 
resulting  in  the  subsequent  field,  and  so  on,  until  the  occurrence  of  an 
act  resulting  in  a  field  which  determines  the  cessation  of  that  variety  of 
endeavor.  In  Lewin's  scheme  of  constructs,  as  in  mine,  the  major  variable 
of  the  internal  situation  (internal  field)  is  some  kind  of  excitation  (with 
direction  and  magnitude) — a  quasi  need,  need-aim,  or  drive;  but  here  I 
am  stressing  the  external  situation. 

It  should  be  noted  in  passing  that  an  adequate  formulation  of  the 
immediate,  or  antecedent,  determinants  of  behavior  can  never  be  given 
in  terms  of  the  instantaneous  external  situation  (configuration  of  space 
or  of  objects  or  of  forces  in  space).  Even  in  the  extreme  case  of  a  wholly 
stationary  external  environment  one  must  take  account  of  the  process 
through  time  of  the  subject's  perception,  apperception,  and  evaluation 
of  the  situation;  and  this  brings  us  to  Whitehead's  actual  occasion,  the 
"real  thing."  In  most  cases,  the  so-called  momentary  external  situation 
(set  of  antecedent  determinants)  is  likely  to  consist,  not  so  much  of  a 
spatial  configuration,  as  of  a  rather  long  pattern  of  symbolic  processes, 
such  as  a  paragraph  of  instructions  read  to  the  subject  by  an  experi- 
menter. But,  let  us  return  to  the  simple  case  of  a  stationary  physical 
field  in  which  a  mobile  person  is  positioned  and  consider  what  kind  of 
map  should  be  made  of  this  so-called  momentary  situation. 

Man  being  a  terrestrial  organism  for  the  most  part — for  the  duration 
of  this  discussion,  anyhow — the  space  to  be  represented  will  be  a  two- 
dimensional  flat  surface,  natural  or  artificial — either  a  circumscribed 
area  of  ground  (composed  of  rock,  soil,  or  sand)  or  a  floor  area  within 
a  building.  This  area  we  shall  call  the  territory  (the  total  spatial  scope 
of  our  concern),  and  this  territory  (say,  a  sparsely  settled  rural  area) 
we  shall  divide  into  regions.,  and  these  regions  into  subregions,  and  so 
on  indefinitely,  if  necessary,  until  we  arrive  at  a  multiplicity  of  places. 
Each  region  will  have  a  certain  area  and  shape  and  will  be  distinguish- 
able from  other  regions  by  the  number,  position,  and  physical  attributes 
(size,  shape,  color,  etc.)  of  its  occupants  (say,  an  assemblage  of  trees,  of 
potato  plants,  of  weeds,  or  of  buildings),  or  by  the  absence  of  occupants, 
and/or  by  boundaries  (walls,  fences,  hedges),  not  to  speak  of  brooks 
and  rivers.  Furthermore,  there  will  be  strips  with  smooth  surfaces  con- 
necting some  of  the  differentiated  regions,  which  I  shall  call  routes,  one 
of  which  will  run  through  a  subregion  occupied  by  buildings,  each 
marked  by  sets  of  symbols,  one  set  indicating  that  food  may  be  purchased 
there,  another  indicating  tools,  another  drugs,  and  another  clothes.  Let 
this  suffice  as  an  account  of  the  structure  of  the  space  relevant  to  our 
problem.  Now  Lewin  was  shrewd  enough  to  see  that  a  map  of  such  a 
territory  showing  the  location  of  physical  objects  and  their  attributes^ 

26  HENRY   A.    MURRAY 

mere  patterns  of  sense  data,  or  mere  primitive  perceptions,  is  of  little 
relevance  to  psychology.  A  modern  artist,  by  a  conscious  effort,  might 
view  his  environment  in  this  way,  or  possibly  a  visitor  from  Mars;  but 
even  in  the  above-given  bare  description  of  the  territory  I  could  not 
without  misunderstanding  omit  such  words  as  trees,  potato  plants, 
buildings,  fences,  routes,  food,  tools,  drugs,  and  clothes,  all  of  which 
words  refer  in  a  rough  way  to  objects  which  not  only  make  themselves 
known  to  our  senses  by  means  of  their  physical  attributes,  but  which, 
under  certain  conditions,  are  capable  of  contributing  to  (or,  in  other 
cases,  subtracting  from)  our  welfare.  Hence,  it  is  not  so  much  the 
physical  attributes  as  such  but  the  known  or  supposed  man-pertinent 
capacities  of  objects  which  influence  behavior  (including  the  capacity 
of  some  objects  to  delight  the  aesthetic  sensibilities  of  the  subject) .  It 
was  these  pertinent  capacities  that  Koffka  and  Lewin  had  in  mind  when 
they  spoke  of  the  "behavioral,"  or  "psychological"  environment,  the 
environment  of  meanings  or  significations. 

This  point  of  view  was  congenial  to  the  one  at  which  I  had  arrived 
with  the  help  of  Uexkiill.  Accustomed  to  the  distinction  between  the 
attributes  and  properties  of  chemical  compounds,  I  had  made  a  com- 
parable distinction  between  what  a  human  object  "looks  like"  and  what 
he  "does"  under  specified  conditions.  Here  I  am  leaving  out,  for  the 
moment,  what  an  alter  does  to  the  subject  solely  by  virtue  of  her  or  his 
physical  attributes  (beauty,  ugliness).  What  an  alter  does,  the  kind  of 
thing  he  does,  to  the  subject.,  I  called  a  press  (plural:  press}.  For  ex- 
ample, the  press  of  Mr.  X  vis-a-vis  a  given  subject  might  be  "to  animate 
him  (the  subject)  intellectually,"  just  as  the  usual  press  of  the  drug 
Benzedrine  when  taken  by  mouth  is  "to  stimulate  mental  processes."  The 
capacity  to  stimulate  is  one  of  the  biochemical  properties  (latent  press) 
of  Benzedrine,  and  when  Benzedrine  passes  into  the  blood  stream  the 
property  becomes  manifest  as  a  process  distinguished  by  its  effect. 
Similarly,  a  known  alter,  regarded  from  the  subject's  point  of  view,  can 
be  represented  as  an  assemblage  of  subject-pertinent  properties,  or  latent 
press,  which  will  be  manifested  as  processional  effects  (operative  press) 
either  spontaneously  or  after  appropriate  stimulation,  when  the  subject 
and  the  alter  meet.  Thus,  as  I  saw  it,  the  physical  structure  of  the  en- 
vironment was  representable  in  terms  of  the  geometric  configuration 
of  regions,  places,  and  objects,  each  with  its  potentially  effective  subject- 
pertinent  properties  (latent  press).  That  strip  of  smooth  surface  over 
there  is  called  a  route  because  from  position  A  to  position  B  it  has  the 
property  of  supporting  a  human  body  or  a  conveyance  and  of  facilitating 
locomotion;  and  boards  cut  from  those  trees  have  properties  suitable  for 
the  excluding  walls  and  supporting  floors  of  houses,  etc.,  etc.  As  a  con- 
sequence of  countless  past  experiences,  such  properties  seem  to  be  re- 

Preparations  for  Scaffold  of  a  Comprehensive  System      27 

vealed  to  us  Immediately  by  mere  perception,  but  at  this  point  I  prefer 
to  speak  of  apperception,  or  apperceptive  perception,  since  it  is  con- 
venient and  often  important  to  distinguish  verbally  betweeil  the  clear 
impression  and  identification  of  a  particular  kind  of  object — say,  a 
hickory  tree — and  the  realization  of  its  properties — say,  the  properties 
of  hickory  which  make  the  wood  especially  suitable  for  ax  handles.  The 
chief  difference  between  the  conceptualization  of  a  pertinent  property 
of  an  inanimate  object,  such  as  a  drug,  and  a  pertinent  property  of  a 
person  (alter)  is  that,  in  the  case  of  the  latter,  one  is  dealing  with  a 
mobile  object  whose  activity  may  be  unprovoked  by  the  subject,  and 
one  must  distinguish  between  an  endeavor  that  fails  (through  incapacity) 
and  an  endeavor  that  succeeds.  It  is  the  difference  between  a  pressive 
disposition  without  ability  and  a  pressive  disposition  with  ability.  But 
more  of  this  later;  I  must  return  to  my  topic. 

There  was  a  wide  gap,  it  always  seemed  to  me,  between  Lewin's 
symbolic  constructs  on  the  level  of  physics  (representative  of  public 
physical  events)  and  his  constant  references  to  a  miscellany  of  wholly 
private  psychic  processes  in  his  subjects  which  he  cleverly  distinguished 
by  intuition,  but  which  he  spoke  about  as  if  they  were  overt  and  obvious 
to  everyone,  or  could  be  reliably  inferred  on  the  basis  of  observed  be- 
haviors. Not  many  psychologists  realized  so  clearly  as  did  Egon  Brunswik, 
that  for  Lewin  the  exterior  field  (the  environment)  was  within  the  sub- 
ject's head.  What  Lewin  called  the  "psychological  environment"  is  the 
subject's  apperceptions  of  the  environment — a  necessary  construct;  but 
it  stood  alone,  no  place  having  been  provided  for  a  more  "objective" 
definition  of  the  environment,  say,  as  apperceived  by  the  psychologist,  by 
selected  judges,  or  by  the  conventional  majority.  Thus  by  Lewin's 
scheme  it  is  not  possible  to  distinguish  between  a  morbid  delusion  and  a 
realistic,  or  congruent,  estimation  of  the  external  situation:  the  situation 
is  exactly  what  the  subject  thinks  it  is,  or  more  accurately — since 
Lewin  rarely,  if  ever,  asked  a  subject — it  is  what  you  think  the  subject 
thinks  it  is  as  you  empathically  perspect  his  thoughts  during  the  course 
of  his  behavior.  Furthermore,  if  the  humanly  pertinent  properties  of  other 
environmental  objects  (as  estimated  by  the  psychologist)  are  never 
mentioned,  we  shall  never  know  how  much  of  the  external  situation  was 
rejected  by  the  subject. 

As  a  step  toward  the  clarification  of  this  issue,  a  number  of  us, 
stimulated  by  an  extended  definition  of  Freud's  important  concept  of 
projection,  conducted  numerous  investigations  of  differences  between  the 
external  situation  as  carefully  and  systematically  perceived  and  apper- 
ceived, say,  by  a  consensus  of  trained  observers  (the  alpha  situation), 
and  the  same  situation  as  perceived  and  apperceived  (under  conditions 
less  favorable  to  accuracy)  by  subjects  with  different  personalities  (each 

28  HENRY   A.    MURRAY 

a  beta  situation),  or  by  subjects  in  a  certain  experimentally  engendered 
temporary  state.  This  is  the  sphere  of  concern  which  is  now  called  "per- 
sonality and  perception/'1 

The  bulk  of  our  experimental  findings  were  unanimous  in  their 
verdict  respecting  the  importance  of  dispositions  (interests,  evaluations, 
and  needful  tensions)  in  determining  the  outcome  of  perceptual,  apper- 
ceptual,  conceptual,  compositional,  and  ordinational  (planning)  pro- 
cesses. In  short,  as  antecedent  determinants  of  overt  behavior,  one  must 
include,  not  only  the  structure  of  properties  and  processes  of  the  con- 
fronting exterior  situations  as  arranged  by  the  experimenter  (cluster 
of  independent  variables,  or  alpha  situation),  but  perceptions  and 
apperceptions  of  certain  of  these  things  (beta  situation)  as  determined 
by  the  dispositional  state  of  a  given  personality  or  type  of  personality 
(cluster  of  intervening,  hypothetical,  or  conventional  variables) . 

Besides  many  other  things,  this  meant  to  me  (with  my  memories  of 
chemistry)  that  a  psychologist  will  bring  in  less  knowledge  by  viewing 
a  person  as  a  mass-point  of  indifferent  constitution  in  a  field  of  forces, 
as  Lewin  (with  his  interest  in  physics  and  his  image  of  Galileo  at  the 
tower)  was  tempted  to  do,  than  he  will  by  viewing  him  as  an  entity 
with  a  particular  conjunction  of  distinguishable  properties.  It  is,  of 
course,  true  that  in  establishing  some  sorts  of  lawful  relationships  be- 

*In  this  and  in  other  related  enterprises,  fortune  favored  me  with  early  col- 
leagues of  the  stature  of  Erik  H.  Erikson,  Donald  W.  MacKinnon,  Saul  Rosen- 
zweig,  R.  Nevitt  Sanford,  and  Robert  W.  White,  of  William  G.  Barrett,  Kenneth 
Diven,  Isabella  V.  Kendig,  Walter  C.  Langer,  Christiana  D.  Morgan,  and  Carl  E. 
Smith;  later  of  Thelma  G.  Alper,  Leo  Bellak,  Vera  V.  French,  Elliott  Jaques, 
Robert  R.  Holt,  Daniel  Horn,  Morris  I.  Stein,  Silvan  Tomkins,  and  Frederick 
Wyatt;  and,  more  recently,  of  Gardner  Lindzey,  of  Anthony  Davids,  Richard  V. 
McCann,  and  Robert  N.  Wilson.  I  have  also  been  advantaged  by  collaborations, 
all  too  short,  with  Freed  Bales,  Tamara  Dembo,  Cora  DuBois,  Walter  Dyk, 
Jerome  D.  Frank,  Christopher  Fried,  Asa  Koht,  Philip  Lichtenberg,  Goodhue 
Livingston,  Charles  C.  McArthur,  H.  Scudder  McKeel,  James  G.  Miller,  Merrill 
Moore,  Hobart  Mowrer,  Benjamin  J.  Murawski,  and  Henry  W.  Riecken,  as  well 
as  with  a  host  of  others  on  the  OSS  assessment  staff  during  the  war  years,  of 
whom  Edward  Tolman  and  John  Gardner  have,  in  interior  dialogues,  admonished 
me  most  often.  Among  these  warm  friends  and  coworkers  I  have  no  reliable  way 
of  apportioning  the  credit  for  leading  me  to  relatively  valid  concepts  and  for 
canceling  many  of  my  least  propitious  errant  speculations,  and  no  reliable  way 
of  apportioning  the  blame  for  withholding  criticism  at  moments  when  I  might 
have  been  deterred  from  this  or  that  cognitive  folly.  Anyhow,  I  am  grateful  for 
the  opportunities  I  have  had  to  serve  as  one  of  many  channels  for  the  ebullient 
ideas  that  have  swirled  and  eddied  round  the  table  at  the  Harvard  Psychological 
Clinic.  And  here  I  must  make  public  my  profound  indebtedness  to  my  good 
friend  and  critic,  Gordon  W.  Allport,  staunch  champion  of  minorities,  without 
whose  timely  advocacy  the  Clinic  might  have  been  dissolved  and  left  no  wrack 

Preparations  for  Scaffold  of  a  Comprehensive  System      29 

tween  entities  it  is  possible  to  disregard  differences  of  constitution,  but 
even  in  physics,  how  often  can  one  predict  the  outcome  of  an  experi- 
ment without  taking  into  account  the  internal  structure  of  the  molecules, 
or  such  properties  of  substances  as  conductivity  or  melting  point?  In 
short,  on  the  down-to-earth  empirical  level  (as  contrasted  with  the 
sphere  of  transcendent,  or  purely  hypothetical,  entities)  one  must  include 
in  one's  formulations  the  properties  (in  specified  states)  of  the  entities 
engaged  in  the  observed  transaction.  For  example,  some  material  entities 
are  nourishing  to  human  organisms,  others  stimulating,  others  soporific, 
and  others  lethal,  and  one  property  of  some  human  organisms  In  a 
certain  (suicidal)  state  is  to  select  a  lethal  rather  than  a  nutritive 
entity  for  incorporation.  I  would  delete  these  references  to  the  obvious, 
if  it  were  not  for  the  fact  that  most  of  us,  in  our  endeavors  to  be  ob- 
jective, to  formulate  behavior  in  terms  of  perceptible  movements — say 
approach  and  ingestion  for  survival — forget,  for  example,  that  poison 
is  attractive  to  persons  in  a  certain  state.  In  short,  we  cannot  throw 
Aristotle  to  the  dogs  and  restrict  our  diet  to  the  more  elegant  formulas  of 
Galileo:  chemistry  is  still  among  the  reputable  sciences  and  closer  to 
psychology — think  of  oxygen,  digestion,  metabolism,  and  endocrines — 
than  is  its  more  admired  older  brother. 

Another  related  conclusion  supported  by  our  findings  was  that  the 
historic  succession  of  the  dispositions  and  experiences  of  a  scientist  has 
a  great  deal  to  do  with  the  concepts  and  theories  that  he  comes  out  with, 
and  largely  because  of  this  conviction,  I  have  often  taken  pains,  as  by 
request  I  am  taking  now,  to  expose  my  inborn  and  acquired  bents  and 
biases,  rather  than  to  make  a  great  to-do  about  my  exemplary  scientific 
objectivity.  It  happens  that  one  of  my  inductions  from  experience  is 
that  many  of  those  who  spend  most  type  asserting  their  immaculate 
empiricism  are  somewhat  below  average  in  their  awareness  of  the 
distorting  operation  of  their  own  preferences  and  ambitions  and,  there- 
fore, are  more  liable  than  others  to  sally  forth  with  reductively  incon- 
gruent  versions  of  reality. 

Additional  concepts  for  the  present  scaffold.  Among  the  other 
conceptual  consequences  of  our  studies  of  personalities  and  their  apper- 
ceptions of  other  personalities  and  of  my  attempts  to  analyze  single 
proceedings,  six  may  merit  definition. 

Cathexis,  From  Freud  I  gratefully  accepted  the  concept  of  cathexis 
(value,  valence)  as  a  useful  variable  in  formulating  personalities  as  well 
as  single  interactions  of  personalities.  But  instead  of  limiting  its  applica- 
tion to  a  loved  person  (the  power  of  an  alter  to  attract,  enchant,  and 
bind  the  affections  of  a  person),  I  defined  it  as  a  possible  disposition- 
evoking  capacity  of  any  kind  of  entity,  or  of  any  kind  of  activity  of  an 
entity,  chiefly  the  capacity  (1)  to  excite  attention  (interest,  concern. 

30  HENRY   A.    MURRAY 

thought,  talk),  or  (2)  to  excite  attention  plus  evaluation,  either  positive 
(favorable — say,  gust,  wonder,  admiration,  love,  approval)  or  negative 
(unfavorable — say,  disgust,  contempt,  disapproval,  distrust,  resentment, 
fear),  or  (3)  to  excite  attention  plus  evaluation  plus  pertinent  activity. 
All  types  of  entities  seem  to  be  capable  of  such  evocation — a  certain 
kind  of  food,  a  homestead,  a  utility,  a  person,  a  social  institution,  a 
novel,  a  moral  code,  a  scientific  theory,  a  philosophy  of  life — and 
similarly  capable  are  all  types  of  activities  of  entities.  Not  only  a  total 
entity,  but  any  part,  integrant,  or  component  activity  of  an  entity  may 
have  the  power  to  attract  attention,  to  please  or  to  displease,  to  instigate 
activity.  You  may  like  a  person  as  a  whole  but  not  like  certain  things 
he  does,  or  you  may  like  certain  things  he  does  but  dislike  him  as  a 
whole.  A  father  spanks  the  boy  he  loves  because  he  hates  lying  and  hopes 
to  spank  this  habit  out  of  him,  and  so  forth.  The  same  might  be  said 
of  the  negatively  cathected  (and  hence  deleted)  parts  of  a  book  in  proc- 
ess of  composition,  a  book  which,  taken  as  a  whole,  may  be  highly 
cathected  by  its  author. 

My  present  notion  of  cathexis  is  not  far  from  the  elaborate  definitions 
of  it  that  were  published  in  The  Clinical  Study  of  Sentiments  [4],  except 
now  the  more  favored  term  is  "value"  and  the  concept  has  been  in- 
corporated in  a  larger  system.  The  term  "sentiment,"  "attitude,"  or 
"established  evaluation"  points  to  dispositional  property  of  a  personality 
which  corresponds  to  the  cathexis  of  an  entity.  One  can  say  that  subject 
A  has  a  strong  sentiment  or  attitude  (established  disposition)  pro  X,  or 
that  his  consistent  evaluation  of  X  is  highly  positive,  or  that  X  has  a 
high  positive  cathexis  or  value  for  A.  Both  terminologies  are  useful.  The 
concept  of  cathexis  is  also  useful,  perhaps  most  useful,  in  indicating  the 
subject's  effect  on  other  people:  in  what  quarters  and  to  what  extent 
he  will  evoke  positive  evaluations,  based,  say,  on  affection,  erotic  love, 
admiration,  or  compassion,  and  leading  to  accessions  or  invitations,  as- 
sociations and  conjugations,  compliances,  services,  or  donations,  etc., 
in  what  quarters  and  to  what  extent  he  will  evoke  negative  evaluations, 
engendered  by  disgust,  contempt,  moral  condemnation,  or  envious 
resentment,  and  leading  to  rejections,  exclusions,  decessions,  expulsions, 
or  inflictions,  etc.  It  is  not  sufficiently  acknowledged,  I  surmise,  that  a 
full  characterization  of  a  personality  should  include,  as  does  the  char- 
acterization of  a  chemical  compound,  the  varieties  of  dispositional  effects 
the  subject  has  on  different  kinds  of  alters. 

Dyadic  system.  The  notion  came  and  stuck  that  a  dyadic  (two- 
person)  relationship,  whether  transient  or  enduring,  should  be  formulated 
as  a  single  system,  equal  analytic  attention  being  devoted  to  each 
participant.  Although  I  have  never  been  inclined  to  accept  Harry  Stack 
Sullivan's  restriction  of  the  domain  of  psychology  to  the  sphere  of  inter- 

Preparations  for  Scaffold  of  a  Comprehensive  System      31 

personal  relations,  I  use  dyadic  interactions  as  a  test  of  every  formulation 
or  theoretical  system  I  encounter  in  the  literature.  If  the  proposed  set 
of  antecedent  environmental  variables  does  not  provide  for  the  definition 
of  an  alter's  subject-oriented  verbal  or  physical  behavior  (e.g.,  such  kinds 
of  "stimulation"  from  the  alter  as  petition  or  command,  praise  or  repri- 
mand, inquiry  or  offering  of  information,  expression  of  good  will,  and  so 
forth),  if  it  does  not  provide  tools  of  this  sort,  then  the  system  is  not 
suited  to  the  representation  of  the  great  bulk  of  human  reactions.  It  may, 
of  course,  have  other  virtues,  but  not  those  I  require:  variables  appro- 
priate to  the  prediction  of  concrete  social  episodes. 

Thema.  The  idea  matured  that  the  basic  pattern  of  a  single  dyadic 
interaction  might  be  most  simply  represented  by  2,  a  symbol  denoting  the 
immediate  direction,  the  need-generated  orientation  (goal),  of  the  pro- 
activity  emanating  from  the  first  interactor,  followed  by  z'z,  a  symbol 
denoting  the  emotional  response  of  the  second  interactor,  and  when  in- 
dicated, a  symbol  denoting  the  need-generated  orientation  (goal)  of  his 
reactivity.  Whether  the  goal  of  the  first  interactor's  (subject's)  activity 
is  the  aim  of  an  independent  need  (and  hence  intrinsically  satisfying  if 
achieved),  or  the  aim  of  a  subneed  (satisfying  if  achieved  although  it  is 
no  more  than  a  subordinate  component  of  a  large  system  of  need-aims), 
or  the  aim  of  a  quasi  need  (merely  instrumental  and  hence  not  in- 
trinsically satisfying)  would  be  a  question  for  further  investigation. 
Months  of  antecedent  study  and  subsequent  exploration  might  be  re- 
quired to  determine  the  probable  status,  or  relative  potency,  of  all  the 
needs  involved  in  a  single  sentence.  The  same  applies  to  the  need- 
determined  response  of  the  alter.  On  this  level  of  formulation  (the 
formulation  of  a  single  proceeding),  it  would  be  sufficient  to  represent 
the  immediate  need-aim  of  the  subject  (proactor)  and  the  need-response 
of  the  alter  f reactor) .  The  need-response  of  the  reactor,  viewed  from  the 
subject's  stamlpoint,  has  been  termed  a  press,  the  alpha  press  being  the 
alter's  actual  response  and  orientation  (in  so  far  as  he  and  the  psy- 
chologist can  define  it)  and  the  beta  press  being  the  subject's  apper- 
ception of  the  alter's  response  and  orientation.  The  simplest  formula, 
then,  would  be  either  an  N-P  (if  the  subject  initiated  the  interaction)  or 
a  P-N  (if  the  alter  acted  first).  This  I  termed  a  simple  micro  thema,  a 
simple  macro  thema  being  an  over-all,  and  hence  much  coarser,  formula- 
tion of  a  longer  transaction,  and  a  serial  thema  being  an  articulated 
procession  of  simple  micro  themas,  which  might  or  might  not  be 
representable  as  a  macro  thema. 

I  might  clarify  this  a  bit  by  illustrating  complementation,  the  simplest 
type  of  dyadic  thema  (others  being  reciprocation,  cooperation,  competi- 
tion, opposition).  Let  us  assume  two  interactors:  X  a  confirmed  trans- 
mitt  or  and  Y  a  confirmed  receptor;  and  then,  out  of  a  large  number  of 

32  HENRY   A.    MURRAY 

complementary  needs,  let  us  choose  the  following  pairs,  and  finally,  let 
us  assume  that  in  each  case  the  episode  is  completed  to  the  satisfaction 
of  both  parties  (criterion  of  a  veritable  complementation) . 

Subject  X,  transmitter  Subject  F,  receptor 

Need  to  inform  (to  relate  facts,  rumors)  Need  for  information  (state  of  interest, 


Need  to  explain  (to  interpret  events)  Need  for  explanation  (state  of  perplexity) 

Need  to  counsel  (to  give  advice)  Need  for  counsel  (state  of  indecision) 

Need  to  amuse  (to  tell  a  funny  story)  Need  for  amusement  (readiness  for  mirth) 

Need  to  console  (to  express  sympathy)  Need  for  consolation  (state  of  distress) 

These  pairs  can  be  taken  to  represent  the  state  of  affairs  in  a  dyadic 
system,  at  the  start  of  five  different  proceedings.  The  transmitter  is 
characterized  by  the  tension  of  a  valued  fullness  (pleni-tension] — he  has 
a  mental  possession  and  the  need  to  impart  it — whereas  the  receptor  is 
characterized  by  lack-tension,  that  is,  by  a  need  for  something,  something 
which,  in  this  case,  the  other  person  is  capable  of  giving  him.  Assume,  for 
example,  X  has  a  mental  possession  (a  funny  story)  which  he  is  keen  to 
communicate  and  Y  is  keen  to  hear  a  funny  story.  As  a  rule,  there  will  be 
mutual  satisfaction  if  the  story  strikes  Y  as  funny  and  he  responds  with  a 
hearty  laugh.  Further  analysis  may  reveal  that  the  apparently  pleni- 
tensive  transmitter  has  nothing  very  interesting  to  say  but  merely  a 
strong  (processional)  disposition  to  babble  (verbosity),  and/or  a  lack- 
tensive  need  for  attention  and  appreciation.  Similar  is  the  next  type  of 
dyadic  pattern,  reciprocation,  except  in  this  case  we  have  a  reciprocal 
complementarity,  the  second  phase  being  marked  by  a  reversal  of  roles — 
the  former  receptor  transmits  with  an  appropriate  degree  of  zest  and  the 
former  transmitter  receives  with  due  appreciation. 

Consideration  of  long  sequences  of  interpersonal  themas  of  this  sort 
has  pretty  much  confused  me  respecting  the  proper  usage  of  the  venerable 
S-R  concept.  The  intended  effect  (need-aim)  of  much  proactive  talk 
(reactive  to  the  mere  sight  of  another  person)  is  an  appropriate  kind 
of  sympathic  response  (press)  from  the  alter  (e.g.,  expression  of  agree- 
ment, compliance,  interest,  mirth,  affection,  admiration,  gratitude,  and 
so  forth),  and  there  seem  to  be  a  good  many  hypomanic  (chemically 
stimulated)  self-starters  and  transmitters  in  the  world  who,  instead  of 
predominantly  responding  to  other  persons,  sail  forth  each  day  full- 
freighted  with,  a  miscellany  of  impatient  stimulations  for  any  ac- 
quaintance (releasor)  who  might  be  capable  of  the  complementary 
responses;  and  when  a  conversation  is  once  launched,  every  response 
is  a  stimulus  to  a  response  which  is  a  stimulus  to  a  further  response,  and 
so  forth,  until  the  tidy  S-R  model  has  been  so  thoroughly  rolled  through 
all  things  that  it  looks  as  if  it  needed  treatment,  some  sort  of  radical 

Preparations  for  Scaffold  of  a  Comprehensive  System      33 

rehabilitation.   Perhaps  it  has  already  been  rehabilitated,  without  my 
knowledge,  by  the  more  advanced  S-R  theorists. 

It  became  evident  in  due  course  that  a  simple  thema,  whether  micro 
or  macro,  is  no  more  than  a  very  coarse,  though  often  meaningful  and 
convenient,  classification  of  an  episode.  To  formulate  an  episode  in  a 
more  refined  way  numerous  other  variables  must  be  included  until  one's 
initially  simple  representation  of  its  major  dynamic  components  has  been 
transformed  into  a  complex  thema.  Among  the  immediate  determinants, 
for  example,  of  Y's  positive  or  negative  reaction  to  a  "funny  story"  told 
by  X,  might  be  the  "appropriateness"  of  the  situation  (never  mind  now 
how  this  is  judged),  the  relative  status  and  degree  of  intimacy  of  X 
and  Y,  the  mirth-potency  of  the  story,  whether  it  is  new  or  stale  to  Y, 
how  well  it  is  told  by  X,  whether  Y  is  momentarily  at  odds  with  X, 
the  current  mood  or  state  of  Y,  the  acuteness  of  Y's  sense  of  humor  in 
general  and  for  this  kind  of  story  in  particular,  how  fastidious  is  his 
standard  of  wit,  to  what  extent  is  Y's  system  of  values  susceptible  to 
offense  by  this  kind  of  story,  and  so  forth.  Just  as  some  psychologists 
have  profitably  devoted  a  professional  lifetime  to  the  study  of  a  hungry 
animal  in  a  maze  containing  food,  so  might  others  spend  rewarding  years 
in  investigating  the  interior  and  exterior  determinants  of  any  one  of  a 
hundred  other  common  types  of  themas,  say,  a  thema  with  an  un- 
successful or  unexpected  outcome,  such  as  "the  joke  that  fails,"  "the 
command  that  is  defied,"  "the  conjugal  proposal  that  is  rejected,"  "the 
injury  that  is  forgiven,"  and  in  each  case,  why? 

My  own  attempts  to  practice  what  I  am  now  preaching — to  explain 
in  some  detail  the  course  of  a  single  type  of  interaction — have  been 
spotty  and  rather  crude,  and,  for  the  most  part,  this  side  of  publication. 
Christopher  Fried,  Philip  Lichtenberg,  and  I  have  separately  spent  two 
years  or  more  investigating  a  few  of  the  determinants  of  the  dyadic 
themas  that  occur  during  film-recorded  competitive  and  cooperate  at- 
tempts to  arrive  at  a  common  plan  of  action;  and,  of  course,  there  have 
been  countless  "clinical"  occasions  for  minute  perceptions  of  other 
common  patterns.  But,  on  the  whole,  the  facts  compel  me  to  acknowl- 
edge that,  except  for  resolute  endeavors  over  the  last  twenty  years  to 
analyze  and  formulate  the  apperceptible  processes  and  products  that 
occur  during  impromptu  compositions  of  dramatic  stories,  I  have  not 
focused  long  enough  on  any  single  type  of  thema  or  on  any  single  method 
of  observation  and  measurement  to  come  out  at  last  with  a  brilliant 
cluster  of  decisive  findings.  Decision  has  been  difficult,  because  if  a 
would-be  personologist  should  elect  to  devote  his  energies  to  the  building 
of  a  miniature  system  of  postulates  and  theorems  applicable  to  the  under- 
standing of  one  kind  of  thematic  unit,  he  would  have  no  time  for  the 
observation  of  other  varieties  of  behavior;  hence  he  would  never  get 

34  HENRY   A.    MURRAY 

around  to  the  practice  of  his  profession,  namely,  the  investigation  of  the 
interrelations  of  the  more  determining  gross  components  of  personality. 

Thematic  dispositions.  It  has  become  more  and  more  apparent  to  me 
that  the  energic  components  of  personality  can  be  better  defined  as 
thematic  dispositions  than  as  general  actional  dispositions.  For  example, 
instead  of  saying  that  X  possesses  the  trait  of  aggressivity,  or  that  he  has 
a  ready  and  intense  need  for  aggression,  one  should,  if  possible,  specify 
the  nature  of  the  pertinent  press  (stimulus)  and  say  with  more  precision 
that  two  of  the  properties  of  his  personality  (I  won't  translate  this  into 
symbolic  shorthand)  are  supersensitive  dispositions  to  react  with  re- 
sentment and  aggressive  words  (1)  to  apperceived  insults  to  his  self- 
respect  and  (2)  to  apperceived  vainglorious  boastings  by  an  alter. 

Serials.  I  was  slow  to  perceive  that  current  psychological  theories  of 
behavior  were  almost  wholly  concerned  with  actions  of  relatively  short 
duration,  reflexes  and  consecutive  instrumental  acts  which  reach  their 
terminus  within  one  experimental  session,  rather  than  with  long-range 
enterprises  which  take  weeks,  months,  or  years  of  effort  to  complete. 
Here,  it  seemed,  was  one  of  the  most  striking  differences  between  men 
and  animals,  namely,  the  capacity  for  time-binding  (Korzybski)  or  the 
span  of  time-perspective  (Frank,  Lewin).  The  behavior  of  animals  can 
be  explained  so  largely  by  reference  to  attractive  or  repellent  presenta- 
tions in  their  immediate  environment  and/or  to  momentarily  urgent  and 
rather  quickly  reducible  states  of  tension;  whereas  a  great  deal  of  a 
man's  behavior  cannot  be  explained  except  by  reference  to  persistent 
"self-stimulation"  in  accordance  with  a  plan  of  action,  which  often  in- 
volves the  subject's  commitment  to  a  distal  goal  or  set  of  goals,  as  well 
as  to  a  more  or  less  flexible  (or  rigid)  temporal  order  (schedule)  or  sub- 
sidiary, or  stage,  goals.  Observing  his  behavior  over  several  months  or 
years,  we  see,  not  only  the  recurrence  of  a  large  number  of  patterns 
devoted  to  the  repetition  of  valued  experiences  and  the  prevention  of 
disvalued  experiences,  patterns  with  homeostatic  effects,  but  a  number 
of  interrupted  successions  of  proceedings  (which  I  am  calling  serials, 
or  long  enterprises),  each  temporal  segment  of  which  is  progressively 
related  to  the  last  (carrying  on  from  where  the  other  stopped),  though 
separated  from  it  by  an  interval  of  time  (commonly  a  day) .  A  successful 
serial  is  different  from  many  day-by-day  reactions  in  so  far  as  its  effects 
are  transtatic  rather  than  homeostatic,  that  is  to  say,  it  transforms  or 
transcends  the  existing  steady  state  by  carrying  a  person  from  one  level 
or  form  of  equilibrium — dispositional,  material,  ideational,  or  social — to 
another:  a  new  interpersonal  relationship  (an  additional  commitment) 
becomes  established;  a  new  house  is  purchased  and  furnished  (which 
must  hereafter  be  kept  up) ;  knowledge  is  gradually  assimilated,  and  a 
new  orientation  (directing  one's  efforts  toward  another  target)  is 

Preparations  for  Scaffold  of  a  Comprehensive  System      35 

acquired;  the  subject  graduates  from  college,  gets  a  job,  and  takes  on  the 
responsibilities  of  a  new  office;  a  novel  is  written  and  published,  and 
so  forth.  Progressive  enterprises  of  this  sort  constitute  the  bulk  of  a 
healthy  young  adult's  endeavors  in  a  "civilized"  society. 

Ordination.  It  took  me  years  to  realize  that  the  psychology  of  the1 
higher  mental  processes  had  been  equivalent,  in  the  niinds  of  most 
psychologists,  to  the  psychology  of  cognition,  and  that  the  psychology  of 
cognition  was  largely  concerned  with  the  processes  whereby  a  person 
acquires  objective  knowledge  and  understanding  of  his  physical  environ- 
ment— the  very  processes  and  the  very  aims  which  are  dominant  in  us 
psychologists — and  that  i,  the  more  fundamental  and  important  knowl- 
edge of  the  satisfying  and  dissatisfying,  the  beneficial  and  the  harmful 
properties  of  the  environment  and  of  the  self  s  capacity  to  cope  with 
them,  and  ii,  the  still  "higher"  mental  processes  involved  in  the  con- 
struction of  a  plan  of  action,  were  pretty  generally  neglected.  What 
should  we  call  the  persistent,  self-critical,  conceptual,  and  often  logical 
mental  processes  that  continue  over  several  months  in  the  mind  of  a 
psychologist  until  they  terminate  with  the  construction  of  an  integrated 
design  for  his  next  experiment?  These  processes  commonly  take  off  from 
perceptions  and  explanations  of  previous  experiments  and  results;  but 
their  immediate  aim  is  not  so  much  to  conceptualize  already  observed 
events  (cognition),  but  to  imagine  something  unobserved — new  condi- 
tions and  new  experimental  operations — and,  by  logic  or  intuition,  to 
predict  the  outcome.  During  his  months  of  planning  the  scientist  (or 
anyone  else  for  that  matter)  is  more  frequently  thinking,  one  might  say, 
on  the  efferent,  rather  than  on  the  afferent,  side  of  the  cortical  arc,  and 
some  psychologists  might,  therefore,  be  disposed  to  subsume  his  mental 
processes  (processes  which  sometimes  occur  very  rapidly — within  a  few 
seconds)  under  conation,  on  the  grounds  that  their  function  is  to  orient 
and  coordinate  action.  But  against  this  is  the  fact  that  they  are  often  very 
"intellectual"  (higher  mental  processes  in  the  strictest  sense),  engaged 
in  a  most  difficult  endeavor  (since  rational  prediction  is  usually  harder 
than  rational  explanation),  and  superordinate  to  other  processes,  in  the 
sense  that  the  goal  and  strategy  which  is  ultimately  selected  will  deter- 
mine behavior  for  a  good  many  months  to  come. 

For  better  or  for  worse  I  have  been  calling  such  mental  processes — 
processes  concerned  with  the  selection  and  integration  of  plans  of  action 
— ordination.  The  preliminary  processes  of  the  imagination — fantasies 
and  trial  experiments  in  the  mind — I  am  calling  prospections.  Here,  in- 
stead of  entertaining  recollections  (replicative  imaginations  of  past 
events) ,  the  subject  is  concerned  with  the  future,  prospectively  picturing 
himself  in  this  and  that  situation,  seeking  this  or  that  opportunity  for 
gratification  or  for  the  advancement  of  his  ambitions.  Here  creativity 

36  HENRY   A.    MURRAY 

may  operate  to  a  marked  degree.  The  selection  from  numerous  alter- 
natives of  a  concrete  and  specific  goal,  purpose,  or  aim  to  appease  one 
or  more  needful  dispositions,  I  am  calling  orientation.  It  is  the  subse- 
quent phase — the  selection  and  temporal  articulation  of  ways-means, 
strategies,  or  tactics  (represented  by  images  or  words) — that  I  am 
calling  ordination.  I  have  found  that  the  word  can  be  used  without 
confusion,  both  for  the  process  of  constructing  (ordinating)  a  plan  and 
for  the  construction  (ordination)  that  results  from  this  process.  An 
ordination  may  have  a  very  short  or  very  long  time  span;  it  may  be 
vague  and  global  or  clearly  differentiated  into  discrete  behavioral  units; 
it  may  be  disjunctive  or  conjunctive  (temporally  integrated  in  a  logical 
manner) ;  it  may  stand  at  any  point  along  the  rigidity-flexibility  con- 
tinuum; and  it  may  have  more  or  less  of  the  property  (power)  of  "im- 
perativeness" (indicated,  partly,  by  shame  or  guilt  if  adherence  to  the 
ordination  is  imperfect) ;  and  so  forth.  One  significance  of  this  concept  is 
its  discrimination  of  a  major  antecedent  determinant  of  behavior  in  a 
"civilized"  society,  namely,  a  fixed  schedule,  the  time  set  for  a  certain 
kind  of  activity,  a  prearranged  appointment,  a  prescribed  order  of  pro- 
cedure— quite  regardless  of  the  mood,  dispositional  state,  need,  or  what 
not,  existing  at  the  moment.  A  good  part  of  socialization  consists  in  ac- 
quiring the  capacity  to  keep  promises,  and  hence,  to  do  something  which, 
at  the  appointed  time,  you  are  not  inclined  to  do.  Furthermore,  we  need 
a  concept  of  prospective  time  reaching  into  an  imagined  future,  some  of 
which  is  filled  (committed,  planned)  and  some  of  which  is  still  unfitted 
(open,  available  for  use) . 


I  came  to  psychology  via  Jung's  Psychological  Types  and  his  Psy- 
chology of  the  Unconscious,  the  first  of  which  initiated  my  professional 
interest  in  types  of  human  nature,  and  the  second,  my  interest  in  uncon- 
scious processes  as  revealed  by  mythologies  and  religious  imagery  as  well 
as  in  the  more  central  and  integral  transformations  of  personality.  What 
I  gained  from  Freud  was  somewhat  more  specific  and  more  applicable 
in  practice  and,  in  due  course,  became  so  much  a  part  of  my  regular 
and  irregular  modes  of  thought  that  there  have  been  times  when  I  forgot 
my  debt  and  took  his  huge  gift  for  granted.  In  the  late  twenties  and 
early  thirities  when  Freud's  name  and  works  were  anathema  to  the 
majority  of  academic  psychologists,  I  was  a  staunch  advocate  and  de- 
fender— as  I  am  now — of  his  greatest  contributions:  (1)  evidences  of 
the  theory  of  unconscious  psychic  processes  and  their  effects,  (2) 
evidences  of  the  determining  importance  of  early  family  relations  and  of 
the  experiences  of  childhood,  of  the  persistence  of  complexes  established 

Preparations  for  Scaffold  of  a  Comprehensive  System      37 

in  those  years,  (3)  countless  illustrations  of  the  multifarious  manifesta- 
tions of  the  sex  drive,  (4)  division  of  the  personality  into  id,  ego,  and 
superego  (conventional  constructs),  (5)  definition  of  several  mecha- 
nisms— repression,  isolation,  denial,  etc. — that  operate  in  the  service  of 
adjustment,  of  self-esteem,  and  of  serenity  of  consciousness,  and  a  host  of 
other  more  restricted  constructs  and  theories  illustrated  by  abundant  case 

I  was  one  of  the  founding  members  of  the  Boston  Psychoanalytic 
Society  and  throughout  the  thirties  was  so  closely  identified  with  its 
cause  that  President  Conant  decided,  primarily  on  these  grounds,  that 
I  was  not  qualified  for  tenure.  Similarly,  in  the  opinion  of  the  psy- 
chologists who  reviewed  it,  Explorations  in  Personality  [3]  was  a  treatise 
out  of  Freud,  or,  more  accurately,  an  attempted  adaptation  of  psycho- 
analytic theory  to  academic  standards.  In  short,  what  I  have  seized  from 
Freud  is  so  very  obvious  that  it  should  not  be  necessary  for  me,  at  this 
late  date,  to  lay  it  on  the  line. 

The  present  situation  is  entirely  different:  Freud  has  conquered.  He 
has  captured  a  large  portion  of  the  Western  mind,  his  revolutionary 
theories  are  learnedly  and  respectfully  discussed  in  General  Education 
courses,  he  is  now  an  indispensable  fixture  in  the  domain  of  psychology, 
and  so  venerated  by  his  professional  disciples  that  his  most  casual  com- 
ments are  repeated  ritualistically  as  absolutes.  Clearly  his  position  is 
assured  and  what  we  all  owe  to  him  is  plain.  The  danger  now  is  precisely 
the  opposite  of  what  it  was  in  the  twenties  when  it  looked  as  if  professors 
were  built  to  shut  their  minds  to  him.  Caught  up  as  we  are  today  in  a 
great  wave  of  Freudiolatry  we  are  inclined  to  take  it  all  as  gospel,  to  feel 
that  the  greater  part  of  what  the  Master  said  is  so  astute  that  the  gestalt 
which  he  created  should  not  be  spoilt  by  calling  attention  to  a  few  trivial 
defects.  This  attitude  would  have  been  impossible  to  Freud  himself  and 
if  continued  its  only  consequence  can  be  sclerosis  of  the  mind  and  rigor 

As  I  weigh  it,  Freud's  contribution  to  man's  conceptualized  knowl- 
edge of  himself  is  the  greatest  since  the  works  of  Aristotle;  but  that  his 
view  of  human  nature  is  exceptionally — perhaps  projectively  and  in- 
evitably— one-sided,  an  extraordinary  abstraction  from  the  abundant 
facts  of  life,  facts  which  may  have  little  bearing  on  the  etiology  of 
neurotic  symptoms  but  great  relevance  to  other  issues.  My  chief  objection 
is  the  commonplace  that  in  his  system,  the  libido  has  digested  all  the 
needs  contributing  to  self-preservation,  self-regard,  and  self-advance- 
ment, together  with  a  host  of  others,  and  rebaptized  them  in  the  name  of 
Sex;  and  that  sex  itself  is  never  given  either  its  profound  evolutionary 
status  or  its  interpersonally  creative  status.  In  the  last  analysis,  it  is 
reduced  to  transient,  superficial,  localized  sensations.  But  then,  who 

38  HENRY  A.    MURRAY 

at  this  preliminary  stage  of  knowledge  can  cover  everything  and  be 


My  Heraclitean  concern  with  process,  change,  and  transformation, 
dating  from  incubator  years  at  the  Rockefeller  Institute,  did  not  gain  the 
impetus  of  a  possession  until,  with  Lucretius  vastly  and  vaguely  in  the 
background,  I  came  upon  Bergson's  theory  of  creative  evolution,  Lloyd 
Morgan's  concept  of  emergence,  Whitehead's  philosophy  of  organism, 
Liebniz's  monad,  and  the  speculations  of  L.  L.  Whyte,  Oparin,  Wald, 
and  others,  respecting  biochemical  evolutions.  What  I  abstracted  from 
these  authors,  in  conjunction  with  a  few  miscellaneous  influxions  from 
the  "unconscious,"  brought  me  to  the  conclusion  that  creativity — the 
formation  of  new  and  consequential  entities  and  of  new  and  conse- 
quential patterns  of  activity — is  a  centrally  determining  capacity  of 
nature,  more  especially  of  human  nature.  I  had  observed  the  progress  of 
morphological  maturations  in  the  embryo  and  later,  the  establishment  of 
new  ordinations  of  serial  enterprises  and  of  new  tactical  patterns  and 
skills  in  personalities;  but  not  until  I  paid  attention  to  analogous  pro- 
ceedings on  the  physicochemical,  sex-genetical,  societal,  and  symbolic- 
representational  levels  and  in  the  sphere  of  technology,  did  I  arrive  at  a 
general  conception  of  formative,  or  constructive,  processes  operating 
throughout  nature. 

What  does  this  amount  to?  First,  a  comprehensive  generalization  re- 
specting a  widely  distributed  capacity  of  entities,  namely,  under  favor- 
able conditions  to  associate  and  remain  associated,  to  combine  and  re- 
main combined,  to  become  involved  in  the  creation  of  new  entities  with 
previously  unexampled  properties,  and  thereby  to  participate  in  the 
making  of  an  irreversible  route  of  events.  Finding  manifestations  of  such 
formative  capacities  at  all  integrative  levels,  we  become  more  assured 
of  their  importance,  more  convinced  that  they  deserve  a  place  in  our 
catalogue  of  fundamental  dispositions.  Also,  we  are  invited  by  the  pos- 
sibility that  detailed  investigations  of  new  productions  at  one  level  may 
suggest  analogies,  correspondences,  and  hypotheses  to  be  tested  at 
another.  Second,  the  observation  that  matter  has  formative  capacities 
makes  us  realize  that  creativity  is  immanent  in  nature,  not  the  pre- 
rogative of  some  transcendent  craftsman,  such  as  Plato's  Demiurge  or 
the  Yahweh  of  Genesis,  nor  imposed  on  nature  by  the  will  of  man.  On 
the  one  hand,  it  permits  a  natural  explanation  of  some  of  the  phenomena 
on  which  the  doctrine  of  vitalism  once  built  its  case,  and,  on  the  other 
hand,  it  shows  us  why  the  term  "mechanism"  (with  its  implicit  reference 

Preparations  for  Scaffold  of  a  Comprehensive  System      39 

to  a  man-made  machine  as  model)  was  not  the  happiest  choice  to 
characterize  the  procession  of  open  systems  under  natural  conditions. 
Our  conviction  that  the  old  vitaBst-mechanist  opposition  is  a  dead  issue 
is  supported,  I  believe,  by  the  abandonment  of  classical  mechanics  by 
physicists  and  chemists  as  basis  for  their  theoretical  inventions.  Third, 
the  addition  of  the  formation  (creation,  construction,  reconstruction) 
process  and  effect  to  our  inventory  of  dispositional  properties  of  per- 
sonality provides  us  with  the  otherwise  missing  necessary  factor  not  only 
for  an  adequate  conception  of  the  liveliest  course  of  mental  processes 
through  time  (the  work  of  the  imagination),  but  for  the  systematic 
representation  of  the  functional  interdependence  of  other  members  of 
the  inventory  during  the  growing,  expanding,  and  developing  phases  of 
a  person's  life. 

The  concept  of  survival  in  one  or  another  guise — self-preservation, 
continuation,  maintenance,  homeostasis,  and  so  forth — can  fulfill  the 
same  construct-integrating  function  in  a  theoretical  system  that  is  de- 
signed to  apply  to  mature  lower  organisms,  since  the  great  majority  of 
their  activities  may  be  partly  understood  historically,  in  terms  of  their 
generally  beneficent  contributions  to  the  continuation  or  the  restoration 
of  a  steady  state.  But  the  principle  of  survival  is  applicable  only  to  the 
status  quo,  not  to  mutation  resulting  in  ontogenetic  and  phylogenetic 
emergent  evolutions.  In  my  view  of  affairs,  anyhow,  it  is  necessary  to 
put  the  processes  of  composition  and  decomposition  at  the  center  of 
things,  between  the  terminus  of  the  afferent  side  and  the  initiation  of 
the  efferent  side  of  the  energy  conversion  arc  of  personality. 

But  this  is  not  the  accepted  view  today — despite  our  great  concern 
with  learning,  with  developments  of  personality,  and,  very  recently,  with 
some  forms  of  mental  creativity.  The  Freudian  inventory  of  drives,  for 
example,  includes  sex,  aggression  (destruction),  and  anxiety-avoidance, 
but  not  construction.  Construction — which,  being  exemplified  on  the 
chemical  level,  is  more  fundamental,  in  my  view,  than  any  of  these 
instincts  as  operationally  defined  by  psychoanalysts — is  subsumed,  in  a 
vague  and  general  way,  under  the  concept  of  sublimation  of  infantile 
eroticisms.  Similarly  in  other  special  fields — sociology  as  well  as  ex- 
perimental psychology.  It  may  be  a  matter  of  time-perspective.  If  we 
are  in  the  habit  of  performing  short  experiments  with  a  peripheral  sub- 
system of  personality,  no  products  of  formative  energies  may  strike  our 
apprehensive  mass;  but  if  we  take  a  longer  view  we  are  struck  by  nothing 

Let  us  assume  a  comfortable  position  on  Ganymede,  satellite  of 
Jupiter,  about  two  billion  years  ago  and  with  supernatural  eyes  take  a 
morning  look  at  the  surface  of  this  planet.  We  shall  perspect,  according 
to  those  who  are  entitled  to  a  guess,  nothing  save  a  fairly  hot  solution 

40  HENRY   A.    MURRAY 

of  inorganic  salts  keeping  company  with  the  simplest  carbon  compounds 
and  enveloping  this  broth  an  atmosphere  of  gases  from  which  oxygen  is 
absent.  In  the  evening  let  us  take  another  look.  Since  we  have  temporarily 
assumed  the  power  of  a  celestial  being,  a  thousand  ages  in  our  sight 
is  as  a  day  gone  by  and  we  shall  now  be  in  the  twentieth  century  gazing, 
I  should  hope  with  wonder,  at  a  tremendous  miscellany  of  natural 
productions — 500,000  kinds  of  organic  compounds,  over  250,000  species 
of  plants,  over  1,000,000  species  of  animals  already  identified  by  man. 
We  shall  perceive  numberless  societal  formations:  human  beings  almost 
everywhere,  behaving  rather  regularly  as  members  of  a  family,  clan, 
tribe,  state,  or  nation,  small  or  great,  with  fairly  consistent  governments, 
laws,  and  policies.  More  obvious  will  be  the  territorial  and  habitational 
constructions :  land  masses  studded  with  settlements,  villages,  towns,  and 
cities,  surrounded  by  cultivated  fields  and  connected  by  paths,  roads, 
boulevards,  and  iron  rails,  running  through  tunnels  and  over  bridges. 
How  long  would  be  a  catalogue  of  man's  material  manufactions, 
architectural,  mechanical,  electrical!  Think  of  the  palaces  and  temples, 
tools  and  armaments,  machines  and  dynamos,  waterworks,  heating 
systems,  lighting  systems,  automobiles  and  airplanes,  and  gadgets  by  the 
millions.  Enough  said.  In  the  name  of  brevity,  let's  skip  the  rest  and 
consider  the  manifold  combinations  of  sounds — the  songs  and  symphonies 
— and  the  combinations  of  images  and  imagined  episodes — the  mythol- 
ogies and  dramas,  sonnets  and  heroic  epics,  histories  and  novels,  and 
their  representations  in  paint,  wood,  and  marble — and  the  combination 
of  concepts  and  reflections — the  ethical  philosophies,  mathematical  for- 
mulations, and  scientific  systems — which  engage  the  minds  of  men,  and 
with  these  let's  end  our  swift  survey  of  entities  and  activities  on  the 
earth's  skin.  All  these  things,  all  varieties  of  social  governments,  material 
conveyances  and  utilities,  symbolisms  and  ideas,  are  productions  of 
the  human  part  of  nature,  and  in  all  probability,  the  vast  majority  of 
them  had  their  genesis  in  the  imaginations  of  a  single  individual  or  of  a 
cluster  of  individuals. 

And  yet,  the  word  "imagination"  has  been  absent  from  the  index 
of  most  textbooks  of  psychology,  and  one  has  to  search  diligently  to  find 
a  little  reference  here  and  there  to  planning  processes  (prospection  and 
ordination),  and  despite  the  emergent  interest  in  creativity,  only  a  few 
authors  have  seen  fit  to  include,  in  some  indefinite  guise  or  other,  a 
formative  disposition — habitational,  implemental,  interpersonal,  social, 
or  symbolic — among  the  properties  of  human  personality. 

Darwin  was  primarily  concerned  with  the  occurrence  of  successively 
more  effective  variations  of  mature  morphologies  from  generation  to 
generation.  In  his  day,  biochemical  science  was  not  so  far  advanced  as 
to  assist  him  with  suggestions  of  plausible  hypotheses  respecting  the 

Preparations  for  Scaffold  of  a  Comprehensive  System      41 

determination  of  these  gross  changes.  Knowing  nothing  of  the  role  of 
chromosomes  and  genes,  of  nucleoproteins  and  DNA,  it  could  hardly 
be  realized  that  chemistry  is  the  instrument  of  heredity.  Today,  however, 
we  can  reasonably  postulate  the  creation  of  new  genes  along  the  route 
of  evolution,  the  mutation  (by  the  transposition  of  a  single  atom  within 
a  molecule)  of  a  gene,  and  a  stupendous  variety  of  possible  combina- 
tions of  genetical  clusters  from  male  and  female.  The  chemists  of 
Darwin's  time  were  not  prepared  to  cope  with  the  problem  of  the 
emergence  of  living  entities  from  nonliving  entities,  the  virus  was  un- 
known; and  the  physicists  were  speculating  about  matters  other  than  the 
possibility  of  the  evolution  of  increasingly  complex  chemical  elements 
and  compounds,  say,  out  of  light  atomic  nuclei.  No  one  had  yet  suggested 
that  as  the  universe  expanded  new  matter  was  constantly  coming  into 
being.  In  psychology,  prevalent  interests  and  conceptions  were  far  from 
the  idea  that  formative  (gestalt-making)  processes  were  involved  in  per- 
ception and  apperception,  not  to  speak  of  their  engagement  in  the 
psychologist's  own  business  of  making  concepts  and  formulating  proposi- 
tions. In  short,  the  data  necessary  for  a  systematic  representation  of 
constructive  processes  on  different  levels  of  integration  were  not  avail- 
able in  the  nineteenth  century.  Today,  however,  a  multiplicity  of  facts 
and  of  reflections  are  at  hand,  enough,  it  seems  to  me,  for  a  rough 
preliminary  draft  of  meaningful  analogies. 

The  very  briefest  outline  I  can  devise,  omitting  several  important 
vectors  and  all  details,  includes  the  movement  (motility,  exploration), 
and  hence,  by  chance,  the  inevitable  contiguity  of  different  entities,  one 
or  each  of  which  is  inherently  attractive  to  the  other — attraction 
(gravitation,  valence,  cathexis)  being  one  of  the  ever-present  forces  of 
the  universe — and,  consequently,  either  symmetrical  or  asymmetrical 
accession  (approach)  resulting  in  an  association  or  structural  formation 
(creation,  construction,  synthesis,  conjugation,  or  incorporation  of  a 
smaller  by  a  larger  entity)  new  to  this  planet,  and  the  cohesion,  the 
sticking  and  staying  power,  and  hence  the  relative  stability  and  longevity 
of  this  unprecedented  form  of  whatever  category — organic  compound, 
genetical  configuration,  family  relationship,  tribal  federation,  govern- 
mental law,  religious  belief,  creed,  or  rite.  If  the  established  form  is  to 
have  further  evolutionary  value  it  must  have  the  attribute  of  plasticity, 
or  flexibility,  the  capacity,  that  is,  to  play  a  part  or  to  become  involved 
in  subsequent  transformations  or  reconstructions.  The  picture  is  one  of 
continuity  through  change.  Only  by  losing  its  particular  identity,  by 
perishing  as  such,  can  a  variation  become  a  link,  stage,  or  episode,  in  an 
evolutionary  sequence,  such  as  the  one  and  only  sequence  that  led  to 
the  human  species. 

Some  of  my  more  earnest  and  literal-minded  friends  remind  me  that  a 


psychologist  should  abjure  fantasies  of  temporal  omniscience  and  keep  off 
of  Ganymede.  Formative  processes  lie  outside  the  sphere  of  psychology : 
they  occur  in  the  "depths,"  behind  the  scenes,  take  a  long  time  to  get 
worked  out,  and  are  wholly  unpredictable.  A  psychologist  should  attend 
to  the  precise  particulars  of  today's  circumscribed  field  of  observation. 
Agreed,  but  suppose  I  ask  one  of  these  friendly  critics  to  serve  as  a 
subject  and  request  him  at  the  first  session  to  demonstrate  his  ability 
to  design  an  experiment  which  will  confirm  or  unconfirm  a  hypothesis 
that  is  unfamiliar  to  him.  In  the  second  session  he  might  be  asked  to 
invent  two  different  parables  to  illustrate  the  evil  effects  of  fanaticism, 
and  in  the  third,  to  outline  a  course  of  action  that  might  happily 
settle  a  specifically  defined  dissension  among  four  members  of  an 
academic  group.  If,  in  each  case,  my  friend  gives  voice  to  the  thoughts 
that  successively  come  to  mind,  the  chances  are  that  we  shall  apperceive 
the  components  of  a  constructive  process  operating  before  our  ears  from 
the  beginning  to  the  end  of  the  experiment — influxions  of  ideas  from 
the  £CwellJ>  of  mind  (What  are  they?  How  fast  do  they  come?  How  varied 
are  they?  How  definite?  How  appropriate  to  the  given  task?),  inter- 
spersed with  evaluations  of  these  influxions,  the  rejection  of  some  and 
the  acceptance  of  others  (How  much  consideration  is  given  to  each 
idea?  How  exacting  is  the  standard  of  assessment?  How  excellent  are  the 
judgments  in  the  opinion  of  experts?  How  much  inhibition,  hesitation, 
censorship,  self-criticism  occurs  along  the  route?  How  quick  are  the 
acceptances?  How  decisive?),  and  then,  to  make  a  long  story  shorter,  the 
temporal  allocations,  or  ordinations,  of  the  accepted  components  of  the 
design,  the  parable,  or  the  plan  (Are  the  concatenations  actually  logical? 
Clearly  expressed?  Have  all  probable  contingencies  been  met?  Has  any- 
thing essential  been  omitted?  How  superficial  or  profound  is  the  offered 
solution  or  composition,  and  so  forth).  In  every  such  experiment  I  sub- 
mit, we  shall  obtain  a  unique  mental  composition  which,  at  one  extreme 
and  perhaps  in  the  majority  of  cases,  may  be  socially  worthless  in  the 
estimation  of  qualified  judges  and  advisedly  forgotten,  but  at  the  other, 
might  be  a  rare  gem  of  creativity,  something  memorable  that  may 
eventually  find  a  place  in  the  great  body  of  cultural  transmissions.  We 
may,  for  instance,  be  dealing  with  a  Whitehead  equal  to  such  utterances 
as  these: 

Insistence  on  clarity  at  all  costs  is  based  on  sheer  superstition  as  to  the 
mode  in  which  human  intelligence  functions. 

No  science  can  be  more  secure  than  the  unconscious  metaphysics  which 
it  tacitly  presupposes. 

Murder  is  a  prerequisite  for  the  absorption  of  biology  into  physics  as 
expressed  in  (its)  traditional  concepts. 

Preparations  for  Scaffold  of  a  Comprehensive  System     43 

A  self-satisfied  rationalism  is  in  effect  a  form  of  anti-rationalism.  It 
means  an  arbitrary  halt  at  a  particular  set  of  abstractions. 

A  science  which  hesitates  to  forget  its  founders  is  lost. 

Scientists  animated  by  the  purpose  of  proving  themselves  purposeless 
constitute  an  interesting  subject  for  study. 

Of  course,  creativity — the  real  thing — is  an  autonomous  and  capri- 
cious process  which  rarely  shows  itself  when  called  upon;  hence,  im- 
promptu tests  are  not  likely  to  bring  forth  anything  but  rather  shallow 
forms  of  originality  and  inventiveness.  Nevertheless,  to  my  way  of  think- 
ing, there  are  compositional  processes  at  work,  ordering  ideas  and 
shaping  sentences — sometimes  brilliantly — in  the  course  of  every  com- 
munication. Most  of  us,  to  be  sure,  make  use  of  the  same  worn  words 
and  trite  phrases  time  and  time  again,  and  integrativeness  in  speech  or 
writing  is  limited  to  the  joining  of  one  commonplace  to  the  next; 
but  were  we  to  abide  by  the  current  laws  of  learning  and  in  talks  with 
friend  or  spouse  repeat  tomorrow  the  response — the  bit  of  news,  the 
joke,  the  idea — that  was  reinforced  today,  we  would  be  heading  for 
press  rejection  or  divorce.  What  we  have  to  learn  is  to  break  a  specific 
speech-reward  connection  and  on  a  subsequent  occasion  substitute  some 
variation.  In  short  we  wiU  be  rewarded  only  for  saying  something  dif- 
ferent from,  but  as  stimulating  as,  that  for  which  we  were  rewarded 
last.  Conclusion:  a  gust  for  novelty  and  emergent  forms  is  widely 
distributed  among  members  of  our  breed. 

For  the  present,  we  may  define  participant  creative  processes  in 
terms  of  their  effect,  result,  achievement,  namely,  an  unprecedented 
form,  and  confine  our  attention  to  stable  forms  which  are  retrospectively 
apperceived  as  valuable  and  as  having  further  consequences  in  an 
evolutionary  context.  Striking  to  many  of  us  is  the  blindness  of  these 
processes,  their  experimental  character,  and  their  resistance  to  the 
coercions  of  conscious  purpose,  which  is  something  that  is  worth  con- 
sidering in  connection  with  human  imaginations,  and  the  occurrence  in 
some  people  of  a  strong  disposition  to  create :  to  combine  sounds,  images, 
words,  concepts,  propositions,  ideas,  ordinances,  people,  things,  strategies, 
or  techniques  in  new  and  significant  forms  which  express  something  that 
is  worth  expressing,  order  things  that  are  worth  ordering,  build  some- 
thing that  is  worth  building,  or  solve  a  problem  that  is  worth  solving. 
Mobilized  by  a  need  of  any  other  class  than  this,  a  human  subject  is 
likely  to  have  a  picture  in  his  mind's  eye  of  what  he  wants — water,  sexual 
intercourse,  a  habitation,  an  automobile,  world  news,  membership  in  a 
certain  group,  promotion,  prestige,  or  what  not.  Under  most  circum- 
stances, what  he  wants  already  exists  somewhere,  actually  or  potentially, 
in  the  environment,  and  he  must  take  it  pretty  much  as  it  is  or  as  it 

44  HENRY   A.    MURRAY 

comes.  There  Is  food  in  that  restaurant,  information  he  requires  in  that 
book,  a  person  over  there  whose  friendship  he  might  win,  a  job  to  be  had 
and  money  to  be  earned,  et  cetera,  et  cetera.  But  the  aim  of  creativity — 
say,  a  design  for  a  more  efficient  machine,  an  architectural  innovation,  a 
symbolic  plot  for  a  drama  to  be  written,  the  explanation  of  an  enigmatic 
phenomenon,  a  more  enlightened  foreign  policy — has  no  existence  any- 
where. A  person  with  this  need  must  work,  think,  brood,  daydream, 
rest,  sleep,  turn  his  thoughts  to  other  things — perhaps  drink  and  read 
detective  stories — until  his  mind  will  favor  him  with  a  representation 
which  possesses,  in  his  prospecting  eye,  the  attributes  that  he  seeks,  and 
then  he  must  be  favored  further  by  representations  of  suitable  embodi- 
ments. A  man  may  rack  his  brains  throughout  a  lifetime  without  re- 
ceiving the  vision  or  idea  for  which  he  longs,  or  if  the  idea  has  come  to 
him,  he  may  labor  for  years  without  finding  the  way  to  expound  it  in 
a  persuasive  manner  or  to  implement  it  in  an  actional  endeavor.  That  is 
to  say,  we  are  dealing  here  with  energies  of  the  human  mind  that  do  not 
respond  directly  to  voluntary  efforts.  Voluntary  efforts  can  influence 
their  direction,  defining,  so  far  as  possible,  the  target  of  their  endeavor, 
but  they  cannot  force  them  to  render  up  the  desired  form  or  answer. 

Nowadays  it  is  pretty  generally  agreed,  I  would  suppose,  that 
imaginations  of  any  real  consequence  are  generated  outside,  or  "below," 
the  stream  of  awareness,  after  a  more  or  less  prolonged  period  of  in- 
cubation, and  they  are  apt  to  leap  to  consciousness  abruptly  at  the  most 
unexpected  moments.  Sometimes,  like  a  dream,  they  seem  to  come  from 
without  rather  than  from  within  the  mind.  A  vision  has  been  called  a 
vision  because  it  is  a  visual  presentation,  a  present,  a  gift,  to  the  inner 
eye,  just  as  the  heavenly  constellations  at  night  are  a  presentation,  or 
gift,  to  the  outer  eye.  It  was  partly  on  this  ground,  we  may  surmise,  that 
the  ancients  believed  that  visions  of  import  came  from  the  gods,  as  best 
among  their  blessings  to  deserving  men.  Today  we  are  disposed  to  say 
that  they  come  from  the  unconscious.  But  the  proposition  I  am  sub- 
mitting here  is  that  the  witting  purpose  to  create  something  with  certain 
valued  properties  is  almost  wholly  blind,  its  goal  being  to  conceive  a  goal ; 
and  though  voluntary  effort  is  one  determinant  of  success,  the  processes 
on  which  creativity  depends  proceed,  for  the  most  part,  spontaneously 
and  autonomously  outside  of  consciousness  and  give  rise  to  hundreds  of 
influxions  which  do  not  survive  because  consciousness  rejects  them,  and 
if  a  certain  influxion  is  considered  worthy  of  survival  it  may  not  be 
what  consciousness  was  seeking,  but  something  else  entirely. 

Facts  of  this  order  constitute  the  basis  for  the  not  uncommon  ex- 
perience among  creative  men  of  serving  as  a  vehicle  or  mouthpiece  of 
some  supernatural  or  superpersonal  imperative,  of  being  an  agent  of 
evolution  instead  of  a  feverish  egoistic  little  self.  "This  is  the  true  joy 

Preparations  for  Scaffold  of  a  Comprehensive  System      45 

in  life,"   Bernard  Shaw  has  written,   "the  being  used  for  a  purpose 
recognized  by  yourself  as  a  mighty  one." 


As  one  of  the  charter  members  of  the  Department  of  Social  Relations 
at  Harvard,  I  could  hardly  fail  to  be  inspired  and  directed  in  my  think- 
ing by  our  largely  shared  ambitious  aim  to  advance  by  successive  trials 
toward  a  common  theoretical  system  for  basic  social  science.  If  it  had  not 
been  for  this  association,  for  the  continuous  influence  of  such  learned 
and  persuasive  colleagues  as  Clyde  Kluckhohn  and  Talcott  Parsons,  I 
might  still  be  representing  personalities  in  so  near  a  social  vacuum  as  we 
did  in  Explorations.  As  a  biologist  I  had  been  attached  to  the  concept 
of  the  herd  instinct,  as  elaborated,  say,  by  Trotter,  and  as  a  psycho- 
analyst, to  the  concept  of  identification  in  each  of  its  different  meanings, 
as  well  as  to  the  several  propositions  respecting  the  internalization  of  the 
parental  superego.  Nothing  is  more  apparent  as  we  look  at  others  and 
ourselves,  especially  in  the  United  States — despite  or  because  of  our 
loudly  avowed  ideology  of  freedom  and  individuality — than  the  tre- 
mendous prevalence  of  unconscious  imitation  and  conformity,  of  the 
educing  and  constraining  force  of  public  opinion  and  behavior.  But  I 
did  not  become  aware  of  the  numerous  cultural  differentiations  one 
had  to  make,  differentiations  of  socioeconomic  classes,  of  special  sub- 
groups, of  rank  in  the  decision-making  hierarchy,  of  role  and  function, 
until  I  gave  a  joint  seminar  with  the  encyclopedic  Kluckhohn,  who 
consented  to  the  office  of  tutor  in  these  matters.  There  I  once  again 
experienced  the  truth  of  the  old  adage :  the  best  way  to  learn  a  subject 
is  to  teach  it,  in  this  case  in  conjunction  with  an  expert.  Besides  my 
indebtedness  to  the  elaborate  classifications  and  generalizations  of  Tal- 
cott Parsons,  I  should  mention  among  other  respected  instructors  in  the 
social  sciences :  Pareto  as  expounded  by  L.  J.  Henderson  in  a  memorable 
seminar,  Malinowski,  Sapir,  Margaret  Mead,  Ralph  Linton,  John 
Dollard,  Florence  Kluckhohn,  Edward  Shils,  Robert  Merton,  Harold 
Laswell,  Ernest  Cassirer — the  list  is  long;  many  congenial  influences 
have  necessarily  been  omitted. 

Since  the  anthropological  and  sociological  concepts  that  I  employ  are 
pretty  nearly  all  derivative,  I  need  not  say  much  on  this  score.  Here 
again  I  have  been  influenced  by  Darwin,  specifically  by  the  theory  that 
the  group  more  than  the  individual  has  been  the  evolutionary  unit.  Being 
of  this  persuasion,  I  have  come  to  think  that  no  theoretical  system  con- 
structed on  the  psychological  level  will  be  adequate  until  it  has  been 
embraced  by  and  intermeshed  with  a  cultural-sociological  system.  Al- 

46  HENRY   A.    MURRAY 

though  every  individual  has  some  measure  of  inner  life,  a  host  of  private 
and  largely  secret  feelings,  fantasies,  beliefs,  and  aspirations,  and  has 
some  extent  of  free  play  outside  the  coercions  and  restraints  of  the  social 
system,  the  great  bulk  of  his  overt  behaviors  are  regulated  by  the  mem- 
berships and  roles  to  which  he  is  committed,  his  actual  behavior  being 
the  resultant  of  a  fusion  or  compromise  between  cultural  specifications 
and  standards  and  his  own  dispositions  and  abilities.  Such  is  the  con- 
ventional doctrine  of  our  time,  in  one  guise  or  another,  and  I  have  little 
to  add  to  it.  But,  since  the  group  theory  of  evolution  is  rarely  mentioned 
today  and  since,  for  better  or  for  worse,  it  has  strongly  influenced  my 
speculations,  I  am  yielding  to  the  temptation  of  quoting  a  few  para- 
graphs from  a  recent  attempt  I  made  to  expound  it  in  a  condensed 

Surveying  the  evidences  of  man's  development  on  earth,  the  later 
Darwin  concluded:  first,  that  the  survival  of  the  fittest  is  a  principle 
which  applies  decisively  not  so  much  to  individuals  as  to  rival  groups — 
tribes,  states,  or  nations — and  second,  that  mutual  sympathy,  aid,  and 
collaboration  among  members  of  a  group  are  conducive  to  its  solidarity, 
and  hence  to  its  combative  power  and  survival.  To  put  it  another  way, 
one  of  the  critical  variations  established  long  ago  was  a  clannish  com- 
bination of  families  more  powerful  than  any  single  person,  a  flexible  yet 
stable  social  system  with  some  differentiation  of  functions  and  conse- 
quently with  an  enhanced  capacity  to  cope  with  various  tasks  and 

From  the  beginning,  if  we  follow  Sir  Arthur  Keith's  composition  of 
the  evidence,  every  successful  group  has  adhered  to  a  double  code  of 
conduct,  a  Janus-faced  morality:  one  face  preaching  submission  to 
authority,  reverence,  cooperation,  loyalty,  good  will,  and  generosity 
within  the  group,  and  the  other  more  contorted  face  shouting  with  rage 
and  murderous  aggression  toward  members  of  opposing  groups.  Other 
things  being  equal,  it  must  have  been  the  clans  or  tribes  which  embodied 
this  dual  standard  in  the  best  balance  that  triumphed  and  endured,  and 
passed  on  to  their  descendants  down  to  the  present  day  the  dispositions 
which  sustained  it. 

This  theory  of  group  evolution  helps  us  to  understand  why  man  is  a 
social,  rather  a  solitary,  self-sufficient  creature  and  why,  as  a  social 
creature,  he  is  both  humane  and  brutal.  Illustrative  of  his  social  prop- 
erties are  such  familiar  facts  as  these :  that  the  vast  majority  of  men  are 
reared  in  one  particular  society,  a  society  that  is  prejudiced  in  its  own 
favor,  and  are  satisfied  to  be  lifelong  interdependent  members  of  this 
society,  that  the  bulk  of  their  enjoyments  come  from  interacting  with 
its  members,  that  they  are  at  peace  with  themselves  only  when  they 

Preparations  for  Scaffold  of  a  Comprehensive  System      47 

feel  and  act  in  accord  with  its  customs  and  ideals,  and  that,  even  in  their 
furthest  reaches  of  self-forwarding  ambition,  they  choose  for  their  most 
delectable  final  prize  the  applause  of  their  fellow-beings,  and  after 
death,  fame,  "that  last  infirmity  of  noble  mind."  The  dual  morality  of 
groups — tribes  and  nations — accounts,  in  some  measure,  for  the  failure, 
the  halfheartedness  and  insincerity,  of  all  attempts  to  abolish  war  and 
for  the  fact  that  human  beings  have  been  generally  so  willing,  even  eager, 
to  suppress  their  fears  of  self-extinction  and  fight  for  their  country  to 
the  tragic  end,  as  well  as  for  the  fact  that  a  man  who  kills  a  hundred 
members  of  an  enemy  society  is  declared  glorious,  but  is  condemned  to 
the  severest  punishment  if  he  stops  the  life  of  a  single  fellow  citizen. 

It  is  supposed  that  the  generally  victorious  groups  were  those  which 
most  fully  incorporated  and  exploited  the  vaingloriousness  and  pride,  the 
greed  and  will  to  power  of  their  individual  members.  But  what  is  the 
significance  of  the  will  to  power?  Power,  intoxicating  as  it  may  be  to 
some  men  and  to  some  nations,  is  a  means  to  something,  not  an  end. 
Power  for  what?  To  this  question  the  response  of  a  creative  evolutionist 
might  be:  power  to  construct  ever  larger  and  less  vulnerable  social 
systems  controlling  ever  larger  areas  of  the  earth's  resources,  or  in  other 
words,  power,  spurred  on  by  greed,  to  grow  and  to  develop,  by  invading, 
conquering,  subjugating,  and  assimilating  weaker  units,  or  more  peace- 
fully and  happily  in  some  cases,  by  federating  with  other  units.  History 
reports  a  great  number  of  such  sequences:  the  integration  of  primal 
groups  into  clans,  and  of  clans  into  tribes,  and  of  tribes  into  small  nations, 
and  the  integration  of  small  nations  into  great  nations  that  subsequently 
broke  apart,  the  rise  and  decline,  the  evolution  and  involution,  of  mighty 
civilizations,  as  Toynbee  has  shown  us,  but  as  yet  no  orchestration  of 
state  sovereignties  into  a  world  order,  no  political  embodiment  of  that 
dream  of  universal  fellowship  which  centuries  of  idealistic  men  have 
recommended  to  our  hearts. 

In  short,  everything  I  have  said  relative  to  formations,  transforma- 
tions, malformations,  and  deteriorations  on  the  psychological  level  is 
applicable  in  a  general,  though  not  specific,  way  to  the  level  of  group 


Although  I  was  educated  on  the  principle  that  limitation  of  aim  is 
the  secret  of  success  in  science,  and  that  the  scientist  is  responsible  for 
particulars,  it  must  be  only  too  apparent  to  you  that  I  have  been  tempted 
to  depart  from  the  wisdom  of  this  strategy  by  the  dream  of  an  all- 
embracing  scheme,  a  unified  science,  not,  of  course,  to  be  achieved  in  my 

48  HENRY   A.    MURRAY 

own  lifetime  but  in  the  distant  future,  if  there  is  to  be  a  future  for  our 

I  suppose  it  would  be  proper  to  speak  of : 

1.  A  comprehensive  concept  (such  as  energy,  process,  matter,  form, 
motion)  which  refers  to  something  that  is  always  and  everywhere  ob- 
servable or  inferable. 

2.  A  comprehensive  conceptual  scheme  (such  as  the  periodic  table, 
classifications  of  botanical  and  zoological  forms)    which  differentiates 
relationally  all  entities  and  all  attributes  and  properties  of  entities  within 
the  domain  of  a  single  discipline. 

3.  A  comprehensive  formulation,  theory,  or  law  (such  as  e  =  me2, 
the  laws  of  thermodynamics,  the  theory  of  evolution)  which  is  applicable 
over  a  wide  range  of  phenomena. 

4.  A  comprehensive  spatial  scope  of  individual  concern  within  a 
single  discipline,  such  as  (to  limit  consideration  to  the  biological  and 
social  sciences)  that  of  a  physiologist  who  takes  the  total  organism  as  his 
province  (rather  than  specializing  in  kidney  function),  that  of  a  psy- 
chologist who  takes  the  whole  personality  (rather  than  specializing  in 
cognition) ,  or  that  of  a  sociologist  who  takes  the  total  community  (rather 
than  specializing  in  family  structure) .  Scope  of  data. 

5.  A  comprehensive  temporal  span  of  individual  concern  within  a 
single  discipline,  such  as  that  of  a  biologist  who  is  interested  in  genetics 
and  heredity,  that  of  a  psychologist  who  is  occupied  with  parental  as 
well  as  subsequent  determinants  of  personality,  or  that  of  a  sociologist 
or  anthropologist  who  studies  historic  transformations.  Span  of  data. 

Now,  one  of  the  best  appraisers  of  the  status  of  psychological  theory 
in  this  country,  the  wisely  chosen  Director  of  this  project,  stated  not  so 
long  ago  that  the  development  of  our  science  had  been  more  retarded 
in  recent  years  by  straining  after  comprehensiveness  than  by  any  other 
variety  of  ambition.  But  since  it  is  not  clear  to  me  which  of  the  above 
forms  of  comprehensiveness  he  had  in  mind,  I  have  not  yet  had  to 
square  my  shoulders  to  the  verdict  guilty.  There  is  at  least  one  form 
of  comprehensiveness  for  which  I  have  not  reached,  the  comprehensive- 
ness of  a  neat  net  of  postulates  and  theorems  that  is  expected  to  catch 
every  kind  of  fish  that  swims  in  the  stream  of  human  experience  and 
behavior.  I  have  never  been  so  optimistic  as  to  think  that  we  psychologists 
were  anywhere  near  the  day  when  some  master  mind  might  achieve  so 
much.  Instead  I  have  been  a  perpetual  catcher  and  collector  of  facts 
and  figures,  a  perpetual  classifier  of  concepts,  and  a  promoter,  in  a  little 
way,  of  marriages  of  concepts,  believing  that  these  pedestrian  occupations 
were  appropriate  to  the  stage  of  conceptual  evolution  at  which  psy- 
chology has  arrived.  Here  I  am  not  speaking  for  the  psychobiologists  who 
study  the  ways-means  learning  processes  of  imprisoned  animals.  They, 

Preparations  for  Scaffold  of  a  Comprehensive  System      49 

so  far  as  I  can  see,  have  already  arrived  at  that  state  of  knowledge  and 
mastery  of  their  variables  from  which  law-making  for  their  territory 
makes  sense. 

The  forms  of  comprehensiveness  of  which  I  am  most  surely  guilty 
are  comprehensiveness  of  territory  in  space  and  time.  I  have  spoken  of 
my  interest  in  creative  evolution  down  the  ages  and  in  developments  of 
personalities  from  birth  on  (temporal  comprehensiveness) ;  and  I  have 
indicated  how  I  was  forced,  to  put  it  bluntly,  by  rny  colleagues  at 
Harvard  to  become  socio-spatially  comprehensive,  concerned  with  the 
supraorganism  of  which  every  personality  is  imperatively  a  functioning 
component.  Nor  can  other  groups,  out-groups  and  foreign  nations,  be 
excluded  from  the  picture,  it  being  all  too  evident  these  days  that  a  little 
shooting  incident  on  some  distant  surface  of  our  planet  might  initiate 
a  global  conflict  which  would  change  the  roles,  the  activities,  and  the 
effects  of  millions  of  human  beings.  Belief  in  the  imminence  of  a 
catastrophic  war  is  currently  one  of  the  determinants  of  anxiety  in  a 
large  number  of  people  occupying  statuses  of  responsibility.  And  then, 
beyond  the  earth  and  its  contentious  nationalities,  revolve  the  sun,  the 
moon,  the  planets,  stars,  and  Milky  Way,  all  of  which  have  influenced 
the  minds  of  countless  individuals  and  collectivities,  not  as  the  Chaldean 
astrologers  surmised,  but  by  drawing  aspirations  and  cognitions  upward, 
by  engendering  images  and  stories  of  celestial  divinities  and  powers,  of 
resurrections  and  ascensions  to  a  heavenly  paradise  beyond  the  grave, 
and  of  life  everlasting  in  a  society  of  musical  winged  beings,  not  to 
speak  of  the  attraction  by  cosmic  bodies  of  astronomers  and  poets. 


From  what  I  have  confessed  so  far  it  must  seem  as  if  the  need  for 
certainty,  powerful  in  most  scientists,  is  very  weak  in  me.  But,  as  I  weigh 
them,  my  hopes  and  expectations  in  this  regard  are  no  higher  and  no 
lower  than  they  legitimately  can  be  nowadays  in  the  sphere  of  endeavor 
to  which  I  am  committed.  Were  my  demands  greater,  I  either  would  be 
perpetually  defeated  or,  to  escape  from  this,  would  be  impelled  to  quit 
personology  and  return  to  chemistry  for  peace.  I  take  heart  from 
Aristotle:  "It  is  the  mark  of  an  educated  man  to  look  for  precision  in 
each  class  of  things.  .  .  .M 

But  this  is  not  the  whole  story.  There  is  something  more  in  me  which 
is  not  irrelevant  to  this  issue:  the  induction  from  experience  that  a 
compulsive  need  for  intellectual  certainty — abetted,  I  would  suppose,  by 
longings  for  personal  security — is  very  apt  to  lead  to  deadly  falsifications 
and  distortions  of  reality.  Leaving  aside  the  changeless  eternal  forms  and 
absolutes  of  philosophers  and  theologians,  and  confining  ourselves  to 

50  HENRY  A.    MURRAY 

scientists,  we  can  find  innumerable  examples  of  the  operations  of  this 
need:  the  selection  of  the  most  fixed,  permanent,  or  recurrent  things 
to  study,  the  unnatural  stabilization  of  the  experimental  environment, 
the  prevention  of  all  but  two  or  three  possibilities  of  response,  the 
circumscription  of  the  area  of  observation  to  a  small  part  of  the  total 
field  of  influential  forces,  et  cetera,  et  cetera.  Such  choices  and  con- 
straints are  valid  parts  of  the  strategy  of  science  and  not  criticizable 
as  such.  They  are  to  be  criticized  only  when  the  results  obtained  in 
highly  focused  studies  of  this  sort  are  generalized  across  the  board  and 
the  notion  propagated  that  the  entities  with  which  we  are  concerned 
are  far  more  structured,  rigid,  stable,  orderly,  consistent,  and  predictable 
than  they  really  are. 

In  my  philosophy  there  are  no  absolute  or  inevitable  laws,  no  en- 
during certainties :  every  observation,  every  inference,  every  explanation, 
and  every  prediction  is  a  matter  of  less  or  greater  probability.  To  this 
most  psychologists,  I  trust,  would  be  ready  to  assent. 


My  interest  in  systems  was  confined  at  first  to  shifts  of  equilibria, 
as  a  function  of  oxygen  tension,  among  the  electrolytes  of  blood  within 
the  walls  of  a  glass  vessel.  The  scope  of  the  next  system  I  studied  with 
some  care  was  a  volume  bounded  by  an  eggshell,  closed  to  material  sub- 
stances but  open  to  intakes  and  outputs  of  gases.  Here  my  chief  source 
of  illumination  was  Elements  of  Physical  Biology  by  Lotka  [2].  But  the 
relevance  of  these  investigations  and  formulations  to  psychology  was  not 
apparent  to  me  until  the  thirties  when  I  was  introduced  to  Pareto's 
representation  of  society  as  a  system,  and  somewhat  later  to  the  con- 
ceptualizations of  the  Chicago  group  as  set  forth,  say,  in  Levels  of  Inte- 
gration in  Biological  and  Social  Systems,  edited  by  Robert  Redfield  [5]. 
Ever  since,  encouraged  by  Whitehead's  speculations,  I  have  been  addicted 
to  the  perilous  practice  of  discovering  analogies  among  events  at  dif- 
ferent levels.  This  hobby,  once  private  and  covert,  has  become  more 
articulate  of  late,  partly  owing  to  parallels  discovered  in  the  writings  of 
L.  von  Bertanlanffy,  A,  E.  Emerson,  R.  W.  Gerard,  and  other  men  who 
are  concerned  with  correspondences  and  differences  between  various 
kinds  of  systems — what  is  now  known  as  General  System  Theory. 

I  am  wary  of  the  word  "system,"  because  social  scientists  use  it  very 
frequently  without  specifying  which  of  several  possible  different  denota- 
tions they  have  in  mind;  but  more  particularly  because,  today,  "system" 
is  a  highly  cathected  term,  loaded  with  prestige;  hence,  we  are  all 
strongly  tempted  to  employ  it  even  when  we  have  nothing  definite  in 
mind  and  its  only  service  is  to  indicate  that  we  subscribe  to  the  general 

Preparations  for  Scaffold  of  a  Comprehensive  System      51 

premise  respecting  the  interdependence  of  things — basic  to  organismic 
theory,  holism,  field  theory,  interactionism,  transactionism,  etc.  For  ex- 
ample, the  terms  "personality-as-a-whole"  and  "personality  system"  have 
been  very  popular  in  recent  years;  but  no  writer,  so  far  as  I  know,  has 
explicitly  defined  the  components  of  a  "whole  personality"  or  of  a 
"system  of  personality."  When  definitions  of  the  units  of  a  system  are 
lacking,  the  term  stands  for  no  more  than  an  article  of  faith,  and  is  mis- 
leading to  boot,  in  so  far  as  it  suggests  a  condition  of  affairs  that  may 
not  actually  exist.  It  suggests  not  only  that  one  is  dealing  with  a  set  of 
recurrent,  orderly,  lawful  interactions,  but  that  the  number,  constitution, 
position,  and  effects  of  the  interacting  units  remain  relatively  constant. 
That  is,  it  is  usually  taken  for  granted  that  "system"  refers  to  a  homeo- 
static,  boundary-maintaining  system.  Finally,  overtones  convey  the  im- 
pression that  the  speaker  has  a  steady,  coherent  theoretical  system  in  his 
head  which  conforms  to  the  steady  coherent  system  he  is  studying. 
Hence  I  am  wary  of  the  word.  But,  having  found  that  I  cannot  get 
along  without  it,  I  must  do  my  best,  when  the  time  comes,  to  define 
my  restricted  usages  of  this  term. 

I  might  say,  in  a  general  way,  that,  for  me,  system  applies  to  a  more 
or  less  uniform  integration  of  reciprocating  and/or  cooperating  func- 
tional activities,  each  of  which,  under  favorable  conditions,  contributes 
to  the  continuation  of  the  entire  cycle  of  activities  which  constitute  the 
system.  As  a  rule,  such  a  system  is  boundary-maintaining.  According  to 
this  view,  each  entity  (form  of  matter)  involved  in  a  cooperating  system 
may  be  called  an  organ,  relative  to  that  system,  each  organ  being  defined 
in  terms  of  process  and  its  contributing  effect,  or  since  organ  processes  are 
not  always  capable  of  achieving  a  contributing  effect,  in  terms  of  their 
direction,  endeavor,  or  intended  effect.  Thus,  each  unified,  boundary- 
maintaining  system  may  be  partially  defined  by  representing  the  integra- 
tion of  successive  processes  and  effects  which  are  required  to  keep  it 
growing  and/or  to  keep  it  going  as  a  unique  and  vital  whole.  The  major 
unitary  functional  systems  with  which  a  social  scientist  is  concerned 
are  these:  personality  systems,  dyadic  social  systems,  polyadic  social 
systems,  representational  (symbolic)  systems,  each  of  which  may  be 
divided — according  to  different  spheres  of  concern — into  large  sub- 
systems. For  example,  a  personality  system  may  be  divided  into : 

1.  A  psychosomatic  system,  consisting  of  all  needs  and  activities 
concerned  with  the  growth  and  welfare  of  the  body:  procurement  and 
incorporation  of  water  and  food,  transposition  and  allocation  of  food 
particles,  differential  construction  of  frame  and  organs,  excorporation 
of  water  and  waste,  actuation  and  integration  of  muscular  patterns, 
development  of  manual  and  athletic  skills,  defense  of  the  integrity  of  the 
body,  etc. 

52  HENRY  A.    MURRAY 

2.  A  psycho-material  system,  consisting  of  all  needs  and  activities 
concerned  with  the  acquisition,  restoration,  or  construction  of  a  territory 
and/or  of  a  habitation  (stead  and  shell),  as  well  as  with  the  acquisition, 
restoration,  or  construction  of  implements  or  machines,  utilization  of 
these  implements,  development  of  technical  skills,  defense  of  property, 

3.  A  psychosexual  system,  consisting  of  all  needs  and  activities  con- 
cerned with  erotic  love :  stimulations  and  interactions,  the  formation  and 
continuation  of  an  erotic  dyad,  conjugations,  and  the  conception  of  off- 
spring, etc. 

4.  A  psychosocial  system,  consisting  of  all  needs  and  activities  con- 
cerned with  nonerotic  social  reciprocations:  transmissions  and  receptions 
of  affection,  of  food,  money,  and  material  entities,  of  information  and 
evaluations,  of  orientations  and  ordinations,  directions  and  compliances, 
development  of  social  skills,  etc. 

5.  A  psycho-representational  system,  consisting  of  all  mental  (cogni- 
tive and  ordinative)   needs  and  activities  associated  with  the  above- 
mentioned  systems — acquisition  of  knowledge,  explanations,  and  postula- 
tions — as  well  as  mental  needs  and  activities  concerned  with  impersonal 
symbolic   systems    (explicit   culture),   with   law,    art,   science,    morals, 
ideology,  and  religion,  development  of  mental  skills. 

The  personality  system,  as  such,  is  concerned  with  the  allocation  of 
time  and  energy  among  these  different  subsystems  and  sub-subsystems, 
the  ordination  of  their  component  serial  endeavors,  the  repression  of  un- 
acceptable emotions  and  impulses,  and  the  reduction  of  conflicts  and 

A  dyadic  system  consists  of  the  interplay  of  two  personality  systems, 
each  of  which  is  given  equivalent  attention.  This  is  enough  to  indicate, 
very  roughly,  the  way  the  term  "system"  is  applied  in  the  scaffold  as 
now  constituted. 


When,  after  finishing  part  1  of  this  assignment — my  autobiography 
of  somewhat  relevant  cerebrations — I  got  round  to  a  closer  examination 
of  the  scheme  provided  us,  I  discovered  that  it  was  even  more  exacting 
than  I  had  initially  believed.  It  was  definitely  beyond  my  reach,  beyond 
the  reach,  I  judged,  of  anyone  who  is  primarily  concerned,  at  this  stage 
of  things,  with  the  formulation  of  different  types  of  personalities  as 
manifested,  say,  by  different  classes  of  reactions  to  a  variety  of  similar 
situations,  rather  than  with  the  reactions  of  most  people,  say,  to  modi- 
fications of  one  particular  experimental  situation. 

I  might  have  profited  by  the  moral  of  the  Icarian  thema,  as  repre- 

Preparations  for  Scaffold  of  a  Comprehensive  System     53 

sented  in  the  careers  of  several  young  persons  assessed  at  the  Baleen,  an- 
nex of  the  Harvard  Psychological  Clinic.  Its  moral  is  that  of  the  inevitable 
fall  of  over-reaching  aspiration,  the  nemesis  of  hubris,  so  familiar  to  the 
Greeks.  But  the  prospect  of  this  outcome  did  not  bring  about  a  reason- 
able abandonment  of  the  project.  It  merely  served  to  check  me  to  the 
point  of  regarding  the  committee's  standard  as  an  unrealizable  ideal, 
but  yet  something  to  be  held  in  view  while  I  labored  over  the  develop- 
ment of  the  scaffold.  As  it  turned  out,  the  effect  of  this  ideal  was  an 
almost  continuous  procession  of  very  general  as  well  as  of  very  particular 
conceptual  compositions,  decompositions,  and  recompositions,  which 
kept  informing  me  of  the  intricate  influence  of  more  and  more  variables 
in  the  determination  of  the  course  and  outcome  of  almost  every  unit  of 
behavior  that  could  interest  a  personologist.  Thus,  I  was  led  on  from 
complication  to  complication,  and  though  many  were  resolvable,  the 
resolutions  served  only  to  increase  the  number  of  aspects  to  be  con- 
sidered and  of  discriminations  to  be  made  in  analyzing,  explaining,  or 
predicting  any  sequence  of  significant  transactions.  After  a  year  or  more 
of  this  sort  of  thing,  the  produce  of  variables  had  reached  an  unmanage- 
able degree  of  refinement  and  of  magnitude;  and,  approaching  the  dead- 
line, I  was  reminded  of  the  judgment  of  Hippocrates:  life  is  short,  the 
art  long,  occasion  instant,  decision  difficult. 


It  is  not  possible  to  pinpoint  in  the  vast  libraries  of  books  and  periodicals 
the  precise  source  of  each  assumption,  concept,  method  that  has  been 
mentioned  in  this  paper.  I  have  included  the  names — all  well-known — of 
the  more  influential  theorists,  but  what  I  have  acquired  from  some  of 
these — Henderson,  Jung,  Prince,  Alexander,  Whitehead,  Lewin,  Kluckhohn, 
Allport,  Parsons,  and  others — came  very  largely  through  conversations  and 
discussions,  and  what  I  have  acquired  from  the  writings  of  these  and  others 
came,  not  from  one  article  or  book,  but  from  pretty  nearly  all  their  works. 
This  is  not  the  place,  it  seems  to  me,  to  list  the  relevant  works  of  Aristotle, 
Darwin,  William  James,  Bergson,  Lloyd  Morgan,  Santayana,  Whitehead,  or 
Gassirer,  or  of  such  social  scientists  as  Pareto,  Parsons,  Lasswell,  Malinowski, 
Sapir,  or  Kluckhohn,  or  of  such  authors  as  Janet,  Freud,  Jung,  Adler,  Rank, 
Alexander,  Horney,  Sullivan,  Kris,  or  Erikson,  or  of  those  psychologists  who 
are  concerned  with  personality,  such  as  McDougall,  Allport,  Murphy, 
Maslow,  Adams,  or  McClelland. 

My  constant  disposition  has  been  to  select  new  fields  of  investigation  and 
to  avoid  those  which  have  already  been  occupied,  if  not  packed,  by  com- 
petent experimentalists.  For  example,  I  have  had  no  first-hand  experience  in 
dealing  with  the  intricate  problems  of  perception  or  of  animal  learning,  and 
hence  I  have  mentioned  but  few  names  of  psychologists  who  have  con- 
tributed to  our  understanding  of  these  phenomena.  Many  of  these  have  in- 

54  HENRY  A.    MURRAY 

fluenced  me  directly  as  well  as  indirectly.  But  it  would  hardly  be  appropriate 
in  this  place  to  list  the  works  of  such  men  as  Pavlov,  Thorndike,  Watson, 
Hull,  Tolman,  Bollard,  Mowrer,  Neal  Miller,  or  Skinner,  or  of  those  who 
have  been  concerned  with  the  philosophy,  logic,  or  semantics  of  theory 
building,  such  authors  as  Bridgman,  Stevens,  Hull,  Lewin,  Koch,  Egon 
Brunswik,  Else  Frankel-Brunswik,  Bergmann,  Meehl,  or  Feigl. 

So  far  as  my  own  bibliography  is  concerned,  the  latest  edition  of  it  can 
be  found  in  C.  S.  Hall  and  G.  Lindzey,  Theories  of  Personality  (1957), 
published  since  the  completion  of  all  that  I  have  written  here. 

The  works  referred  to  in  the  text  are  these: 

1.  Henderson,  L.  J.  The  order  of  nature.  Cambridge,  Mass.:  Harvard 
Univer.  Press,  1917. 

2.  Lotka,  A.  J.  Elements  of  physical  biology.  Baltimore,  Md.:  Williams  & 
Wilkins,  1925. 

3.  Murray,  H.  A.,  et  al.  Explorations  in  personality.  New  York:  Oxford 
Univer.  Press,  1938. 

4.  Murray,  H.  A.,  &  Morgan,  Christiana  D.  A  clinical  study  of  senti- 
ments.  Published  separately  and  in  Genet.  Psychol.  Monogr.,   1945,  No. 

5.  Redfieldj   R.    (Ed.)    Levels   of  integration   in    biological  and  social 
systems.  Lancaster,  Pa.:  Jaques  Cattell,  1942. 



Austen  Riggs  Center 

Formerly  I  found  it  extraordinarily  difficult  to  accustom  my  readers  to 
the  distinction  between  the  manifest  dream-content  and  the  latent  dream- 
thoughts.  Over  and  over  again  arguments  and  objections  were  adduced  from 
the  uninterpreted  dream  as  it  was  retained  in  the  memory,  and  the  necessity 
of  interpreting  the  dream  was  ignored.  But  now,  when  the  analysts  have 
at  least  become  reconciled  to  substituting  for  the  manifest  dream  its  meaning 
as  found  by  interpretation,  many  of  them  are  guilty  of  another  mistake,  to 
which  they  adhere  just  as  stubbornly.  They  look  for  the  essence  of  the 
dream  in  this  latent  content,  and  thereby  overlook  the  distinction  between 
latent  dream-thoughts  and  the  dream-work.  The  dream  is  fundamentally 
nothing  more  than  a  special  form  of  our  thinking,  which  is  made  possible 
by  the  conditions  of  the  sleeping  state.  It  is  the  dream-work  which  produces 
this  form,  and  it  alone  is  the  essence  of  dreaming — the  only  explanation  of 
its  singularity  [S.  Freud:  The  Interpretation  of  Dreams,  98,  pp.  466-^-67]. 

Introduction 57 

I.  Background  Factors  and  Orienting  Attitudes  {!}*.      .......  59 

A.  Background  Factors 59 

B.  Orienting  Attitudes 63 

1 .  The  nature  and  limits  of  psychological  prediction 63 

2.  Level  of  analysis 66 

3.  Utility  and  role  of  models 67 

a.  The  reflex-arc  (or  topographic)  model 67 

b.  The  entropy  (  or  economic)  model 68 

c.  The  Darwinian  (or  genetic)  model 68 

d.  The  Jacksonian  (or  neural  integration  hierarchy)  model  ....  70 

e.  A  combined  model 71 

'The  completion  of  this  study  was  aided  by  the  Ford  Foundation's  grant  in 
support  of  research  at  the  Riggs  Center. 

*  The  bracketed  numbers,  when  they  occur  in  the  tables  of  contents  of  the  essays  in 
this  volume,  indicate  items  in  the  Suggested  Discussion  Topics  relevant  to  the  headings 
which  they  follow.  See  Note  on  the  Use  of  Discussion  Topic  Index  Numbers  in  the  Appendix. 



4.  The  comprehensiveness  of  empirical  reference 78 

5.  Quantification  and  mensuration 79 

6.  Formal  organization 82 

II.  The  Structure  of  the  System  {2} 82 

A.  The  Subject  Matter  of  Psychoanalysis  Is  Behavior  (the  Empirical  Point 
ofView) 82 

B.  Behavior  Is  Integrated  and  Indivisible:  the  Concepts  Constructed  for  Its 
Explanation  Pertain  to  Different  Components  of  Behavior  and  Not  to 
Different  Behaviors  (the  Gestalt  Point  of  View) 83 

C.  No  Behavior  Stands  in  Isolation:  All  Behavior  Is  That  of  the  Integral  and 
Indivisible  Personality  (the  Organismic  Point  of  View) 85 

D.  All  Behavior  Is  Part  of  a  Genetic  Series,  and  through  Its  Antecedents,  Part 
of  the  Temporal  Sequences  Which  Brought  About  the  Present  Form  of  the 
Personality  (the  Genetic  Point  of  View) 86 

E.  The  Crucial  Determinants  of  Behaviors  Are  Unconscious  (the  Topo- 
graphic Point  of  View) 88 

F.  All  Behavior  Is  Ultimately  Drive  Determined  (the  Dynamic  Point  of 
View) 89 

G.  All  Behavior  Disposes  of  and  Is  Regulated  by  Psychological  Energy  (the 
Economic  Point  of  View) 91 

H.  All  Behavior  Has  Structural  Determiners  (the  Structural  Point  of  View)  93 

I.  All  Behavior  Is  Determined  by  Reality  (the  Adaptive  Point  of  View)  .  97 

J,  All  Behavior  Is  Socially  Determined  (the  Psychosocial  Point  of  View)  .  101 

K.  Discussion 104 

III.  The  Initial  Evidential  Grounds  for  the  Assumptions  of  the  System  and 
Their  Strategic  Character  {3} 110 

A.  Initial  Evidential  Grounds 110 

1 .  The  assumption  of  psychological  determinism Ill 

2.  The  assumption  of  unconscious  psychological  processes 112 

3.  The  assumption  of  unconscious  psychological  forces  and  conflicts    .      .  112 

4.  The  assumption  of  psychological  energies  and  their  drive  origin      .      .  113 

B.  Strategic  Choice  of  Initial  Evidential  Grounds 114 

C.  The  Relation  of  the  Observations  to  the  Theory 116 

IV.  Construction  of  Function  Forms  {4} 121 

V.  The  Problem  of  Quantification  {5} 124 

A.  Cathexes 125 

B.  Dimensional  Quantification 129 

VI.  The  Formal  Organization  of  the  System  {6} 133 

A.  The  Present  Status  of  the  System 133 

B.  The  Desirable  Level  of  Formalization 135 

VII.  The  Range  of  the  System's  Applications  {7} 136 

VIII.  History  of  the  System's  Research  Mediation  {8} 138 

IX.  The  Evidence  for  the  System  {9> 140 

The  Structure  of  Psychoanalytic  Theory  57 

A.  Current  Status  of  Positive  Evidence 140 

B.  Major  Sources  of  Incompatible  Data 143 

C.  "Critical"  Tests  of  Principal  Assumptions 148 

X.  Methods,  Concepts,  and  Principles  of  Broad  Application  {10}    .      .      .      .  149 

A.  The  Range  of  Application 149 

B.  Methods,  Concepts,  and  Principles  of  Long-term  Significance      .      .      .  150 

1.  Methods 151 

2.  Principles 152 

3.  Concepts 153 

a.  Dynamic  point  of  view 153 

b.  Economic  point  of  view 153 

c.  Structural  point  of  view 1 53 

d.  Genetic  point  of  view 154 

<?.  Adaptive  point  of  view ....  1 54 

XI .  The  Theory's  Achievements  and  Its  Convergence  with  Other  Theories  {1 1}  155 

A.  Achievements 155 

B.  Convergence  with  Other  Theories 157 

XII.  Tasks  for  the  Future  Development  of  the  Theory  {12} 159 

A.  Empirical  Evidence  Needed 159 

B.  Obstacles  to  the  Development  of  the  Theory 161 

C.  The  Practical  Obstacles  to  Theoretical  Advance  in  Psychology        .      .  163 
References 167 


Neither  Freud's  nor  other  psychoanalysts'  writings  give  a  systematic 
statement  of  the  psychoanalytic  theory.  This  fact,  combined  with  my 
acceptance  of  the  outline  suggested  by  Dr.  Koch  (reflected  in  my 
section  headings),  imposed  problems  that  the  writers  of  the  other  essays 
may  not  have  had  to  face.  It  is  proper,  therefore,  to  state  the  premises 
of  this  essay. 

1.  Freud's  writings  are  the  source  of  psychoanalysis  and  provide  the 
frame  of  reference  for  its  systematic  treatment.  Thus  this  essay  centers 
on  Freud's  work. 

2.  A  systematic  treatment  of  the  theory  should  also  take  into  account 
other   contributions  which  decisively  shaped  the  present  form  of  the 
theory.  Thus  this  essay  draws  extensively  on  Hartmann's  and  Erikson's 

3.  The  systematic  statement  of  the  theory  should  establish  its  rela- 
tion to  the  alternative  ("Neo-Freudian")  theories  which  arose  from  it. 
But  an  early  attempt  at  systematization,  such  as  the  present  one,  can 
neglect  them  without  prejudice.  Thus  this  essay  barely  touches  on  Adler, 
Jung,  Rank,  Homey,  Kardiner,  and  Sullivan. 


4.  An  attempt  at  systematization  should  stay  as  close  as  possible 
to  the  formulations  of  the  literature,  but  it  should  also  interpret  these. 
This  essay,  therefore,  although  it  hews  close  to  the  existing  theory,  does 
draw  inferences  and  does  make  judgments.   Consequently  the  theory 
it  presents  may  appear  unfamiliar  to  the  reader  whose  conversance  with 
psychoanalysis  is  exclusively  clinical  or  only  cursory. 

5.  A  systematic   statement  need  not   follow  the   emphases  of  the 
literature.  Its  emphasis  should  be  dictated  by  systematic  considerations. 
Thus  this  essay  only  touches  on  the  theory  of  symptoms,  psychosexual 
development,    therapy    (e.g.,    transference   and   resistance),    and    con- 
centrates heavily  on  what  Freud  called  metapsychology.   It  makes  a 
distinction  between  what  might  be  called  the  special  or  clinical  theory 
and  the  general  or  psychological  theory  of  psychoanalysis. 

6.  One  of  the  aims  of  Dr.  Koch's  outline  was  to  make  the  essays  of 
these  volumes  comparable.  To  fulfill  this  requirement  I  found  it  necessary 
to  present  some  considerations  (e.g.,  on  independent,  intervening,  and 
dependent  variables,  as  well  as  on  quantification)  which  have  no  direct 
roots  in  the  psychoanalytic  literature  and  which  enter  frames  of  ref- 
erence somewhat  alien  to  my  own  thinking. 

7.  Dr.  Koch  suggested  that  the  authors  of  these  essays  assume  the 
reader's  familiarity  with  previous  statements  of  the  theory  and  dwell 
primarily  on  systematic  issues.  Complying  with  his  outline  made  some- 
thing of  this  sort  unavoidable.  Yet  I  had  to  conclude,  from  recent  writ- 
ings of  psychologists  about  psychoanalysis,  that  familiarity  with  the  psy- 
chological theory  of  psychoanalysis  (as  distinguished  from  the  psycho- 
analytic theory  of  neurosis)  cannot  be  generally  assumed.  The  historical 
relationships — which  play  an  important  role  in  all  unsystematized  the- 
ories— seem  to  be  particularly  unfamiliar.  Thus,  time  and  again,  I  found 
it  necessary  to  summarize  theories  and  to  sketch  historical  relationships. 
The  result  of  my  attempt  to  reconcile  these  conflicting  demands  is  not  a 
happy  one.  In  the  beginning  of  the  essay  the  reader  will  find  familiarity 
with  many  concepts  and  theories  taken  for  granted,  only  to  encounter 
some  of  them  later  on,  again  and  again  discussed  in  detail,  with  further 
information  added  each  time.  The  time  limitation — unavoidable  in  such 
collective  endeavors  as  these  volumes — permitted  me  no  better  solution ; 
it  is  also  responsible  for  the  length  of  the  essay.  Had  I  prepared  it  on  my 
own  schedule,  it  would  have  matured  for  a  few  more  years,  and  it  might 
have  become  more  comprehensive  and  tighter  in  its  structure  and  "log- 
ical joints.53 

To  my  mind  it  is  too  early  to  attempt  a  systematization  of  the 
psychoanalytic  theory.  A  science  can  be  a  "good  science"  without  being 
ready  for  a  systematic  presentation:  all  old  sciences  were  once  in  this 
position.  The  existence  of  this  essay  is  thus  in  need  of  explanation.  I  was 

The  Structure  of  Psychoanalytic  Theory  59 

prompted  to  write  it  partly  by  the  urgings  of  Drs.  Gill,  Hartmann,  Holt, 
Klein,  and  last  but  not  least,  Dr.  Koch,  the  coordinator  of  this  APA 
project,  and  partly  by  my  wish  to  pave  the  way  for  an  adequate 
systematic  presentation  of  psychoanalysis. 

The  very  prematurity  of  this  attempt  had  curious  consequences.  The 
essay  presents  several  cross  sections  (for  instance,  models,  points  of 
view)  of  the  theory  which,  though  they  are  linked  by  identical  concepts 
and  by  common  empirical  referents,  are  not  systematically  related  to 
each  other.  The  clearest  indication  of  prematurity  is  the  uncertainty 
whether  we  are  not  yet  able  to  connect  these  systematically,  or  whether 
they  need  not  or  cannot  be  connected. 

Since  the  literature  directly  bearing  on  the  system  of  psychoanalytic 
theory  is  meager,  I  refer — contrary  to  custom — to  mimeographed  ma- 
terial of  limited  circulation  and  even  to  unpublished  manuscripts.  The 
English  Standard  Edition  of  Freud's  writings  is  still  incomplete,  therefore 
the  references  are  to  that  medley  of  editions  which  I  have  used  over  the 
years  in  my  studies.  Some  of  these  involve  inaccuracies  corrected  by  the 
Standard  Edition.  While  I  am  aware  of  these,  I  did  not  attempt  to  cor- 
rect them. 

The  contradictions  between  this  survey  and  the  Rapaport-Gill  study, 
which  went  to  press  since  this  manuscript  was  prepared,  are  explained 
partly  by  the  survey  character  of  this  study  and  partly  by  the  time  lag. 

As  much  as  space  permitted,  I  have  referred  to  sources  and  acknowl- 
edged the  specific  help  I  received.  Drs.  M.  M.  Gill,  R.  R.  Holt,  G.  S. 
Klein,  and  R.  Schafer  read  the  manuscript,  and  their  suggestions  and 
corrections  were  so  numerous  that  without  a  heavy  addition  of  footnotes, 
this  is  the  only  way  I  can  acknowledge  my  indebtedness  to  them.  To  Dr. 
Holt  I  am  particularly  grateful,  not  only  for  his  repeated  readings, 
suggestions,  and  criticisms,  but  also  for  the  share  he  had  in  shaping 
the  considerations  on  variables  and  quantification.  But  I  am  in  even 
greater  debt  to  Erik  Erikson,  Merton  Gill,  Heinz  Hartmann,  and  Samu 
Rapaport.  Last  but  not  least,  I  want  to  express  my  gratitude  to  Mrs. 
Ruth  Shippey,  Mrs.  Barbara  Kiley,  and  Miss  Suzette  Annin.  Mrs. 
Shippey  and  Mrs.  Kiley  did  the  secretarial  work  on  the  several  versions 
of  this  manuscript,  and  Miss  Annin  did  the  editorial  and  bibliographic 


A.  Background  Factors 

The  formative  influences  in  Freud's  background  were  the  Jewish 
tradition,  an  early  developed  interest  in  literature  (particularly  a  de- 
votion to  Goethe  and,  through  him,  to  ancient  Rome),  courses  with 


Brentano  of  act-psychology  fame,  the  impact  of  Darwin's  theory  of 
evolution,  clinical  and  laboratory  research  in  neurology  and  neuro- 
anatomy  (in  the  orbit  of  men  from  Helmholtz's  circle),  clinical  psy- 
chiatric work  (with  Meynert),  clinical  work  with  neuroses  (at  first  with 
Breuer,  Charcot,  and  Bernheim),  and  self-observation.2 

The  influence  of  Helmholtz  on  Freud's  theory  is  seen  in  the  postulate 
of  thoroughgoing  determinism,  in  the  central  position  of  the  pleasure- 
pain  principle  (and  the  primary  process)  which  is  patterned  on  the 
concept  of  entropy,3  in  the  reality  principle  (and  the  secondary  process) 
which  is  patterned  on  the  principle  of  least  action,  and  in  the  "economic 
principle"  which  is  patterned  on  the  principle  of  conservation. 

The  experience  in  neurological  research  is  responsible  for  Freud's 
conception  (derived  from  Hughlings  Jackson's  view  of  the  nervous 
system)  of  a  series  of  psychological  organizations  (instances,  structures) 
hierarchically  and  topographically  superimposed  upon  each  other.  That 
experience  is  also  responsible  for  the  conception  of  associative  networks 
organized  superficially  by  contiguity  but  fundamentally  by  drives,  for  the 
conceptions  of  inhibition  and  facilitation,  at  first  bodily  transported  into 
his  system  from  neurology,  and  for  his  early  assumption  that  psycho- 
dynamics  is  neurodynamics.  Even  when  abandoned,  this  assumption  still 
lingered  on  in  the  form  of  the  belief  that  sooner  or  later  psychodynamics 
would  be  placed  on  the  "solid  footing"  of  neuro-  and/or  biochemical- 

But  Freud's  laboratory  research  was  also  closely  related  to  the  theory 
of  evolution,  and  it  is  probably  this  conjunction  which  is  reflected  in  the 
genetic  cast  of  Freud's  thinking,  particularly  in  the  close  relation  hypo- 
thesized between  phylogenesis  and  ontogenesis,4  in  the  emphasis  on 
epigenesis,  in  the  regression  concept  and  many  others.  A  Neo-Lamarckian 
version  of  evolution  theory  also  seems  to  have  influenced  Freud's 
thinking  [115,  p.  64]. 

The  effects  of  his  clinical  psychiatric  experience  with  Meynert  and 
his  related  readings  (e.g.,  Greisinger),  though  probably  crucial,  have 
not  been  studied  in  detail.5  It  seems  reasonably  certain,  however,  that 
the  contents  of  the  hallucinations  in  "Meynert's  amentia"  served  as  the 
prototype  for  the  concept  of  "wish-fulfillment"  [cf.  35,  p.  136;  98,  pp. 

2  This  list  represents  a  narrow  view  of  "formative  influences."  For  a  broader, 
more  psychological  one,  see  Erikson  [63,  64,  65]  and  Gross  [154];  see  also  [15,  16, 
17,  18;  193;  309]. 

3  Freud  refers  it  to  Fechner  [69,  sec.  11,  p.  94,  note];  see  [123,  pp.  3,  4]. 

4  Dr.  F.  Schmidl  (Seattle)  calls  attention  to  Haeckel's  particular  influence. 

5  But  see  Hartmann's  recent  study  [165]. 

The  Structure  of  Psychoanalytic  Theory  61 

509,  533]  and  as  the  foundation  for  what  will  be  described  below 
as  the  primary  model  of  cognition.  It  is  also  probable  that  the  often 
claimed  influence  of  Herbart  [see  193]  and  the  less  noticed  one  of 
Hering6  on  Freud  came  indirectly  through  the  work  with  Meynert  and 
related  reading.  Neither  his  neurological  research  (dissection  and  micro- 
scopy) nor  his  clinical  psychiatric  work  provided  Freud  with  experience 
in  the  experimental  method:  both  fostered  his  bent  toward  observation. 

The  experience  with  neurotics  left  its  trace  on  Freud's  system  in  the 
recognition  that  psychopathological  phenomena,  such  as  'the  unana- 
tomical  delineation  of  hysterical  symptoms,  are  organized  on  principles 
other  than  those  familiar  to  neuroanatomy  and  neurophysiology;  in  the 
recognition  of  the  power  of  psychological  forces  (through  the  observation 
of  hypnotic  and  waking  suggestion  with  Breuer  and  Bernheim) ;  in  the 
recognition  of  the  existence  of  nonconscious  psychic  formations  (through 
observations  of  hypnosis  and  alternating  states  of  consciousness  by 
Charcot  and  Janet),  and  the  conceptualization  of  these  as  the  System 
Unconscious;  and  in  the  recognition  of  the  crucial  role  of  sexuality  in 
neuroses  (Freud  attributed  his  first  inkling  of  this  to  Charcot's,  Chrobak's, 
and  Breuer's  incidental  comments) . 

The  influence  of  self-observation  (including  his  self-analysis)  is 
ubiquitous  in  Freud's  theory,  and  accounts  for  the  method  of  free  as- 
sociation, for  the  role  of  dream  interpretation  as  an  investigatory  tool, 
and  for  many  specific  discoveries. 

The  traces  of  Brentano's  act  psychology7  are  less  obvious  and  have 
never  been  explicitly  discussed.8  Yet  the  central  position  of  instinctual 
drives  in  Freud's  theory  parallels  Brentano's  interpretation  (which  con- 
trasts sharply  with  that  of  Anglo-Saxon  empiricists)  of  both  stimulation 
and  response  in  terms  of  acts  of  intending.  In  the  early  phases  of  Freud's 
ego  psychology,  Brentano's  influence  seems  even  more  striking.  The  term 
intention  crops  up,  the  problem  of  reality  testing  leads  to  an  analysis  of 
the  "belief  in  reality"  [119,  p.  146]  along  Brentano-like  lines,  and  the 
distinctions  between  what  is  perceived  and  what  is  conceived,  what  is 
real  and  what  is  only  thought,  etc.,  come  into  play.  This  influence 

fl  Ernst  Kris  (personal  communication,  Jan.  11,  1957):  "I  have  noted  one  of 
the  most  obvious  sources  for  Freud's  thinking,  namely  Hering's  paper  on  memory. 
The  evidence  of  Freud's  interest  reaching  up  to  1922  is  absolutely  conclusive  and 
as  far  as  I  know  never  noticed.  It  might  amuse  you  to  look  in  this  connection 
at  Anna  Freud's  translation  of  Levine's  book  on  the  Unconscious.  The  translation 
of  the  chapter  on  Butler  is  by  Freud  and  so  is  an  interesting  footnote." 

7  Concerning  Freud's  contacts  with  Brentano,  see  Merlan  [232,  233]. 

8  Dr.  F.  Schmidl  suggests  that  it  was  through  Brentano  that  Freud  came  to 
know  of  Maudsley,  to  whom  he  refers  in  The  Interpretation  of  Dreams,  and  whose 
concept  of  the  unconscious  may  have  influenced  Freud's. 


pervades  the  Papers  on  Metapsychology  [108,  110,  114,  115,  116,  117, 
119,  120].  And  although  Freud  deliberately  refused  to  have  anything  to 
do  with  philosophy,  he  did  acquire  some  familiarity  with  it  through 
Brentano.  In  one  of  the  few  specific  references  he  makes  to  philosophy, 
he  characterized  psychoanalysis  (and  particularly  its  concept  of  un- 
conscious determination)  as  the  psychological  counterpart  of  Kant's 
philosophical  views  [117,  p.  104].  Indeed,  the  epistemological  implica- 
tions of  psychoanalysis  are  closest  to  Kant  [see  262]  and  most  remote 
from  Anglo-Saxon  empiricism. 

The  influence  of  literature  in  general  and  Goethe  in  particular  are 
again  hard  to  trace.  They  certainly  shaped  Freud's  interest  in  and  grasp 
of  human  nature.  They  provided  the  pattern  for  the  case  history  as  a 
tool,  which  medical  case  histories  of  his  time  did  not  supply  (compare 
the  best  of  these,  Charcot's,  for  example,  to  Freud's).  Indeed,  it  might 
be  said  that  the  intrinsic  validity9  of  his  reasoning  and  descriptive  writing 
often  had  to  serve  him  as  that  indicator  of  validity  which  in  older  sciences 
is  usually  provided  by  quantitative  measures.  He  became  one  of  the  out- 
standing masters  of  thought  and  pen  in  the  German  language  (Goethe 
Prize).  These  influences  also  fostered  in  him  that  sensitivity  to  the 
subtleties  of  verbal  communication  and  that  readiness  to  seek  meaning 
behind  meaning  which,  combined  with  a  knack  for  metaphor  and 
symbol,  are  the  requisites  of  interpretation.  Indeed,  they  probably  guided 
him  to  his  central  conceptions — motivations,  affects,  and  conflicts — 
which  are  the  raw  material  of  all  art. 

The  role  of  the  Jewish  tradition  in  Freud's  thinking,  methods,  and 
theorizing  has  not  been  explored  in  detail  either.  Wittels  [327],  Reik 
[286],  and  Erikson  [63,  64,  65]  have  elucidated  some  aspects  of  it.  It 
is  possible  that  much  of  what  we  attribute  to  Freud's  interest  in  literature 
comes  from  the  tradition  of  "the  people  of  The  Book."  Associative  and 
interpretive  methods  have  some  of  their  most  striking  archetypes  in  the 
methods  of  the  Talmud.  The  stereotyped  Aramaic  phrase,  introducing 
Talmudic  interpretation,  translated  into  English  reads:  "What  does  he 
want  to  let  me  hear?"  But  the  degree  of  Freud's  direct  conversance  with 
his  Jewish  tradition  and  its  effect  on  his  thinking  have  not  yet  been 

^9By  intrinsic  validity  I  mean  what  literary  criticism  means  when  it  speaks  of 
a  "valid  statement" :  the  great  writer  achieves  a  form  which  makes  the  expression 
of  his  observations,  feelings,  and  thoughts  a  "valid  statement."  But  even  in  every- 
day life,  some  of  us  convey  an  experience  so  that  it  is  clear,  convincing,  and 
pregnant  with  meaning,  while  the  reports  of  others  are  pale,  pointless,  and  diffuse, 
as  if  they  were  third-hand. 

9a  In  the  period  of  reading  proof  of  this  article,  I  noted  advertisements  for  an 
apparently  pertinent  publication  (D.  Bakan.  Sigmund  Freud  and  the  Jewish 
Mystical  Tradition.  Princeton:  Van  Nostrand). 

The  Structure  of  Psychoanalytic  Theory  63 

We  cannot  abandon  this  survey  of  background  factors  without  point- 
ing to  one  whose  significance  has  so  far  not  been  explored,  the  Zeitgeist 
[see  31,  32].  Darwin  and  Helmholtz  were  certainly  part  of  it.  Freud's 
Jewish  background  and  fate,  and  their  influence  on  his  thinking,  have 
never  been  discussed  in  the  broad  terms  of  Zeitgeist.  Only  Erikson10  has 
discussed  the  influence  of  the  general  atmosphere  of  Victorian  Vienna, 
which  was  probably  as  much  a  limiter  of  Freud's  social  psychology  as  a 
source  of  his  emphasis  on  sexuality  and  its  vicissitudes.  The  assessment 
of  the  Zeitgeist  attains  particular  urgency  because  of  the  fundamental 
similarities  of  Freud's  theory  to  one  immediately  preceding  it,  and  one 
immediately  following  it  in  time.  Marx,  Freud,  and  Einstein,  who  con- 
tinued the  Copernican,  Kantian,  and  Darwinian  revolutions,  relativized 
our  conceptions  of  the  world.  Marx,  reversing  Hegel's  dictum,  asserted 
that  "man's  [economic]  existence  determines  his  consciousness  and  not 
his  consciousness  his  existence,"  and  thus  made  man's  view  of  his  world 
relative  to  his  socioeconomic  status.  More  broadly,  Freud  asserted  that 
man's  view  of  and  relation  to  his  world  are  dependent  upon  (relative 
to)  his  impulses  and  are  not  simply  imprinted  on  him  by  his  experience. 
Most  broadly,  Einstein  asserted  that  observation  is  relative  to  the  ob- 
server's position.  If  it  should  turn  out  that  the  commonality  of  the  three 
theories  is  as  real  as  it  seems,  and  is  rooted  in  the  Zeitgeist,  then  we 
would  have  before  us  a  background  factor  which,  though  subtle  and 
nonspecific,  might  prove  the  most  pervasive  and  most  powerful  of  all. 

B.  Orienting  Attitudes 

1.  The  nature  and  limits  of  psychological  prediction.  Prediction  in 
psychology  implies  the  postulation  of  thoroughgoing  determinism  in 
human  behavior.  Freud's  assumption  of  exceptionless  psychological 
determinism,  which  is  perhaps  too  easily  taken  for  granted  today,  pro- 
vides the  necessary  foundation  for  prediction. 

Since  the  empirical  material  first  dealt  with  was  the  already  present 
neurotic  symptoms,  Freud's  primary  causal  problem  was  postdiction 
rather  than  prediction.  This  initial  situation  is  not  unique  to  psycho- 
analysis. It  has  its  counterparts  in  the  social  sciences,  e.g.,  in  history, 
and  in  the  natural  sciences,  e.g.,  in  the  theory  of  evolution.  A  theory  is 
not  invalidated  by  being  postdictive,  as  long  as  postdiction  is  carefully 
distinguished  from  ex  post  facto  explanation. 

Because  the  observations  were  made  in  the  therapeutic  situation,  the 
predictions  were  of  necessity  related  to  the  effects  of  therapeutic  inter- 
ventions and  thus  were  fraught  with  the  same  difficulties  which  have 

"Erikson  [64]  also  calls  attention  to  the  influence  that  the  economic  theories 
of  the  time  seem  to  have  had  on  Freud's  thinking. 


beset  recent  investigations  involving  "participant  observation"  and 
"action  research."  Freud  believed  that  only  first  hand  experience  with 
the  psychoanalytic  method  of  treatment  could  make  understanding  and 
valid  tests  of  psychoanalytic  predictions  possible,  that  the  psychoanalytic 
theory  could  be  validated  only  by  the  psychoanalytic  method,  and  that  it 
had,  indeed,  been  so  validated  and  needed  no  further  validation.  The 
method,  the  theory,  and  its  validation  were  considered  inextricably  tied 
to  one  another.11 

While  these  limitations  on  verification  and  prediction  were  pro- 
fessed, from  the  beginning  observations  of  hypnotic  phenomena  were  in- 
voked to  validate  the  propositions  concerning  the  unconscious  [19,  35]. 
Hypnotically  induced  dreams  [306],  hypnagogic  [310]  and  daydream 
phenomena  [319]  were  hailed  as  independent  evidence  corroborating 
the  predictive  (and  postdictive)  power  of  the  theory.  Moreover,  eth- 
nology [109],  literature  [102],  and  psychotic  products  [107]  were  in- 
creasingly invoked  as  independent  confirming  evidence,  though  their  use 
to  corroborate  the  theory  tended  to  merge  with  the  use  of  the  theory  for 
their  explanation.12 

More  recently,  it  became  increasingly  evident  that  direct  studies  of 
infant  and  small-child  behavior  were  needed  for  the  independent  con- 
firmation of  the  theory's  postdictive  reconstructions  of  these  phases  of 
life,  and  many  such  investigations  were  undertaken.  Psychodiagnostic 
and  experimental  evidence  has  also  been  increasingly  invoked  as  con- 
firmation of  the  theory,  though  the  investigations  by  which  this  evidence 
has  been  obtained  have  rarely  shown  due  regard  for  the  complexity  of 

11  The  discussion  of  "critical  tests,"  in  Section  IX.C.  below,  points  out  that 
whereas  in  other  sciences  tests  validating  a  theory  decide  between  alternative  and 
mutually  exclusive  possibilities,  as  a  rule  this  is  not  possible  for  psychoanalytic 
theory.  The  alternatives  envisaged  by  psychoanalysis  are  not  mutually  exclusive 
but  rather  equivalents  which  can  substitute  for  each  other,  according  to  the  dy- 
namics of  the  situation.  Thus  the  theory  is  not  built  by  tests  of  predictions  ex- 
cluding all  but  one  of  several  alternatives,  but  rather  by  the   inclusion  of  all 
observed  alternatives  which  are  consistent  with  the  existing  theory.  Only  those 
alternatives  which  clash  with  the  existing  theory  are  excluded.  The  observation 
which   suggests   such   incompatible   alternatives   is   rechecked   by   further   clinical 
observations.  Rechecks  which  confirm  the  incompatible  alternatives,  and  thus  do 
not  permit  their  exclusion,  lead  to  the  modification  of  the  theory.  It  is  thus  that 
postdiction — guided  by  the  aim  of  preserving  the  internal  consistency  of  the  theory, 
rather  than  by  the  principle  of  parsimony — becomes  the  principal  means  of  theory- 
building  in  psychoanalysis. 

12  All  sciences  must  subject  observations  to  interpretation  in  order  to  establish 
their  evidential  significance  for  the  theory.  This  is  particularly  conspicuous  in 
psychoanalysis,  where  the  concepts  are  by  and  large  at  a  considerable  distance 
from  the  observations.  For  a  further  discussion  of  this  point,  see  pp.  116  fL 

The  Structure  of  Psychoanalytic  Theory  65 

the  theory,  and  their  methods  have  rarely  been  such  that  the  results 
could  pass  as  confirmations  of  the  theory.13 

Psychoanalysis  did  concern  itself  with  one  class  of  predictions, 
namely,  prognoses.  These  are  part  of  the  subject  matter  of  applied 
psychoanalysis  (clinical  psychoanalysis  and  psychiatry)  and  not  of 
theoretical  psychoanalysis  proper.  The  problem  of  prognosis  has  three 
facets:  the  prognosis  for  treatment  by  the  psychoanalytic  method,  the 
prognosis  for  "spontaneous  remission,"  and  the  prognosis  for  treatment 
by  modified  psychoanalysis  or  other  therapy.  So  far  the  study  of  the 
criteria  of  prognosis  has  yielded  rules  of  thumb  rather  than  theory, 
yet  the  concepts  of  "ego  strength"  [158],  "model  technique"  and  "param- 
eters of  technique"  [5 1]  did  arise  in  this  context 

Once  the  postdictive  character  of  psychoanalytic  propositions  is  clear, 
another  characteristic  of  the  theory  also  becomes  obvious.  The  detailed 
study  of  dreams,  of  symbolism,  of  slips  of  the  tongue,  of  wit,  of  as- 
sociation sequences,  and  the  like,  suggests  that  psychoanalysis  studies 
and  predicts  behavior  on  this  "microscopic"  level;  yet  the  actual  aim 
of  the  theory  was  always  to  predict  or  postdict  life-sized  ("macroscopic") 
segments  and  sequences  of  behavior.14  This  curious  duality  is  char- 
acteristic of  the  theory:  it  is  holistic,  but  not  because  it  lacks  methods  for 
studying  and  predicting  the  "microscopic";  and  it  is  atomistic  in  the 
sense  that  it  can  and  does  study  the  "microanatomy"  of  behavior,  but  not 
because  its  methods  and  interests  limit  it  to  "microscopic"  phenomena. 
Naturally,  the  verification  of  its  theory  of  slips  of  the  tongue  by  post- 
hypnotic  suggestions  [53]  or  the  verification  of  its  theory  of  symbolism 
by  means  of  suggested  dreams  [306,  288,  244,  83],  which  involve  "micro- 
scopic" predictions,  does  not  verify  the  "macroscopic"  relationships  pre- 
dicted by  the  theory;  in  turn,  verification  of  macroscopic  relationships 
(e.g.,  that  of  homosexuality  to  paranoia  [cf.  246])  does  not  necessarily 
confirm  the  detailed  mechanisms  (such  as  projection)  which,  according 
to  the  theory,  mediate  these  macroscopic  relationships. 

In  conclusion :  the  nature  of  the  material  Freud  worked  on  led  him 
to  overemphasize  postdiction  and  underemphasize  prediction  in  building 
his  theory.  In  this  he  was  also  influenced  both  by  the  type  of  neurological 
work  he  and  his  teachers  pursued,  and  by  the  methods  of  the  biological 
science  of  the  time.  But  it  may  be  questioned  whether  or  not  any  science 
in  its  beginnings  has  been  free  from  such  imbalances.  The  basic  necessary 

13  The  trouble  with  these  investigations  is  that  either  their  status  as  a  source  of 
independent  evidence  for  the  theory  is  not  established,  or  their  relevance  to  the 
theory  is  not  established.  Cf.  Section  V.,  below. 

14  The  terms  "microscopic"  and  "macroscopic"  are  used  here  in  the  sense  indi- 
cated by  the  examples,  without  reference  to  any  of  their  various  usages  in  the 


condition  for  predictions  and  for  their  confirmation  is  present  in  the  theory 
of  psychoanalysis,  and  certain  types  of  psychoanalytic  predictions  have 
been  confirmed.  Moreover,  postdiction,  if  properly  handled,15  is  as  valid 
a  confirmation  of  a  theory  as  prediction.  The  task  ahead  is  to  add  to  the 
necessary  conditions  of  prediction  the  sufficient  conditions^  by  tighten- 
ing the  theory  and  by  developing  adequate  methods  of  quantification  and 

2.  Level  of  analysis.  The  level  of  analysis  has  changed  repeatedly  in 
the  history  of  psychoanalytic  theory. 

First,  Freud  (1895)  made  an  attempt  [94]  to  account  for  all  be- 
havior by  neurodynamics,  though  even  in  this  period  he  already  had  a 
clear  outline  [35]  of  his  psychological  theory,  which  centered  on  the 
conflict  between  environment  and  ego  (memory  of  traumatic  experience 
vs.  social  propriety  and  self-respect).  At  this  point,  he  equated  the  ego 
with  consciousness  (i.e.,  the  dominant  ideational  complex)  and  the  un- 
conscious with  what  the  environment  disapproved  of.  Thus,  early  psycho- 
analysis operated  with  three  "levels  of  analysis":  neuroanatomy  and 
neurodynamics,  environment  vs.  ego,  Conscious  vs.  Unconscious. 

Second,  in  the  next  phase  (1900)  of  the  theory  [98],  "intrapsychic 
dynamics,"  centering  on  the  drive  vs.  censorship  conflict,  becomes  the 
causal  referent  of  all  behavior  and  the  ultimate  causal  factor.  But  even 
in  this  period  censorship  and  secondary  process  are  connected  by  Freud 
with  reality  and  interpersonal  relations  (environmental  and  psychosocial 
referents).  Yet  the  dominant  level  of  analysis  is  the  intrapsychic  one,  in 
terms  of  drives  vs.  censorship. 

Third,  with  the  development  of  ego  psychology  (1923),  a  dual  intra- 
psychic reference  system  crystallizes  [126]:  drives  and  structures  are 
juxtaposed.  The  dominant  level  of  analysis  is  still  the  intrapsychic  one,  in 
terms  of  drives  vs.  structures. 

Fourth,  (1926)  the  structural  concepts  are  recognized  in  part  as 
representing  external  reality  referents17  and  the  drives  are  recognized  as 
representing  biological  referents.18  Thus  the  intrapsychic  reference  system 

15 The  difficulties  in  confirming  postdictions  are  these:  the  data  on  which  a 
postdiction  is  based  must  in  some  inferable  form  imply  the  relationships  to  be 
postdicted;  however,  the  relationships  implied  in  the  data  must  not  be  so  obvious 
as  to  make  postdiction  superfluous.  "Proper  handling"  of  postdiction  thus  has  to 
make  explicit  both  what  is  given  in  the  data  on  which  the  postdiction  is  to  be 
based  and  what  is  not  given  and  can  be  only  inferred  by  postdiction.  This  is 
easier  said  than  done,  however. 

13  See  Benjamin  [1 1]  for  the  first  discussion  of  this  issue  in  the  literature. 
See  also  footnote  20. 

17  Cf.  Section  II.  H.,  below. 

ts  Freud  wrote:  "  .  .  .  'instinct'  appears  to  us  as  a  borderland  concept  between 
the  mental  and  the  physical,  being  both  the  mental  representative  of  the  stimuli 

The  Structure  of  Psychoanalytic  Theory  67 

is  reduced  to  organism  vs.  external  reality,  and  a  variety  of  hypothetical 
constructs  (drives  and  structures)  are  interpolated.  There  are  now  three 
levels  of  analysis :  biologic,  intrapsychic,  and  real,  though  all  of  these  are 
handled  in  terms  of  their  psychological  representations. 

Fifth,  (1937-1946)  the  psychosocial  referents  crystallize  in  the 
work  of  Horney  [181,  182],  Kardiner  [194,  195],  and  Sullivan  [313] 
on  the  one  hand,  and  in  that  of  Erikson  [56,  57,  59,  60]  and  Hartmann 
[157]  on  the  other.  A  system  of  multiple  levels  of  analysis  evolves,  in- 
cluding the  dynamic,  economic,  structural,  genetic,  and  adaptive  levels, 
whose  foundations  had  already  been  built  in  the  earlier  phases. 

In  conclusion:  the  psychoanalytic  theory,  by  its  conception  of  "over- 
determination,"  kept  itself  open  to  all  relevant  "levels  of  analysis,"  and 
was  not  limited  to  a  single  one  as  were  many  other  theories.  Yet  the 
"intrapsychic"  concepts  in  general,  and  the  drives  in  particular,  remain 
central  to  the  theory. 

3.  Utility  and  role  of  models.  Freud's  theory  contains  four  distinct 
models.  They  are  united  in  the  theory  itself,  but  not  in  one  single  model. 
We  will  first  present  each  of  these,  and  then  attempt  to  develop  a  com- 
bined model. 

a.  The  reflex-arc  (or  topographic]  model.  This  model  [98,  pp.  498ff.] 
represents — as  it  does  in  the  stimulus-response  theories,  too — the  tend- 
ency of  the  organism  to  respond  to  stimulation.  The  Freudian  model, 
however,  has  additional  specifications : 

1.  This  tendency  is  regarded  as  a  direction  of  psychological  processes. 

2.  It  is  one  of  the  two  directions  excitations  can  take,  the  other  being 
the  regressive. 

3.  In  the  ideal  case  the  excitation  begins  in  a  sensory  stimulation, 
passes  through  the  Systems  Unconscious,  Preconscious,  and  Conscious, 
and  terminates  in  motor  action:  this  is  the  "topographic"  course. 

4.  Not  every  excitation,  however,  need  pass  through  the  complete 
topographic  sequence. 

For  instance,  excitations  can  originate  in  the  Unconscious:  drive- 
excitations  usually  do  so,  though  drive  action  is  often  triggered  by  a 
stimulus.  Excitations  can  also  originate  in  the  Preconscious :  dreams  are 
initiated  by  preconscious  day-residues.  Nor  is  it  necessary  that  excitations 
initiated  by  a  sensory  stimulus  run  the  whole  topographic  course;  they 
may  terminate,  temporarily  at  least,  in  the  Unconscious  or  Preconscious : 
that  this  is  the  case  with  "unconscious"  and  "preconscious"  perceptions, 
which  are  clinical  commonplaces,  has  been  confirmed  by  the  experi- 
ments of  Poetzl  [257],  and  others  [e.g.,  Huston,  Shakow,  and  Erickson. 

emanating  from  within  the  organism  and  penetrating  to  the  mind,  and  at  the  same 
time  a  measure  of  the  demand  made  upon  the  energy  of  the  latter  in  consequence 
of  its  connection  with  the  former"  [115,  p.  64]. 


188;  Diven,  47;  Fisher,  84;  Klein  et  al.,  201].  Similarly,  an  excitation 
can  terminate  in  consciousness  without  initiating  a  motor  response. 
Recent  developments  in  psychoanalytic  ego  psychology  demonstrate  that 
autonomous  functions  of  the  ego  (particularly  automatized  ones)  may 
short-circuit  the  topographic  course.  Thus  a  place  within  psychoanalytic 
theory  is  allocated  to  automatized  (one-to-one)  stimulus-response  rela- 
tions. The  topographic  locus  of  origin  is  an  important  characteristic  of 
excitation  processes. 

This  model  was  useful  in  two  ways.  On  the  one  hand  it  coordinated 
descriptively  a  welter  of  otherwise  disparate  observations,  such  as  the 
vicissitudes  of  stimulations,  the  alternative  (ideational,  affective,  action, 
and  abeyant)  responses  to  stimulations,  the  lack  of  one-to-one  relation- 
ships between  stimuli  and  responses,  and  the  wide  variety  of  apparently 
"spontaneous"  ideational,  affective,  or  action  responses  (ranging  from 
dreams,  daydreams,  delusions,  blushing,  sweating  to  parapraxes  and 
random  movements) .  On  the  other  hand,  it  served  as  the  foundation 
for  the  topographic  point  of  view  in  general,  and  for  the  concepts  of  the 
Systems  Unconscious,  Preconscious,  and  Conscious  in  particular,  and 
these  in  turn  were  the  predecessors  of  the  structural  point  of  view. 

b.  The  entropy  (or  economic]  model.  This  model  [98,  pp.  509,  533] 
— implicit  in  the  direction  attributed  to  the  course  of  excitation  in  the 
topographic  model — is  the  crucial,  topographically  incomplete19  sequence 
of  infant  behavior:    restlessness  — >  sucking  on  the  breast  ->  subsidence 
of  restlessness.  This  sequence,  which  makes  behavior  the  referent  of 
tension-reduction  processes,  is  regarded  as  the  basic  model  of  all  moti- 
vated behavior,  and — in  keeping  with  the  postulate  of  determinism — 
pertains  to  obviously  motivated  behaviors  as  well  as  to  apparently  ac- 
cidental ones.  It  can   be  modified — as  we  shall  see — to  account  for 
tension-maintaining  and  tension-increasing  processes  also.  The  merit  of 
this  model  is  that  it  coordinates  a  wide  range  of  phenomena,  and  serves 
as  the  foundation  for  the  concepts  of  the  pleasure  principle  and  wish- 
fulfillment  in  particular,  and  the  economic  point  of  view  subsuming  them 
in  general.  It  plays  an  important  role  in  the  transformation  of  the  topo- 
graphic into  the  structural  point  of  view,  and  also  contains  the  core  of 
the  dynamic  and  adaptive  points  of  view.  Since  this  model  already  im- 
plies some  of  the  others,  we  will  later  present  a  sketch  of  a  previous 
attempt  [267]  to  develop  it  into  a  unified  psychological  model  of  psy- 
choanalytic theory. 

c.  The  Darwinian  (or  genetic]  model.  This  model  [101;  cf.  also  1], 
which  asserts  that  the  course  of  ontogeny  abides  by  inborn  laws,  served 
Freud  as  the  frame  of  reference  for  systematizing  the  data  of  his  patients' 

19  See  p.  67,  above. 

The  Structure  of  Psychoanalytic  Theory  69 

life  histories,  and  became  the  foundation  for  the  genetic  point  of  view  in 
general,  and  for  the  theory  of  psychosexual  (libido)  development,  in- 
cluding the  concepts  of  fixation  and  regression  in  particular.  Freud's 
inclination  to  alloy  the  Darwinian  model  with  Haeckel's  biogenetic  law 
(ontogeny  repeats  phytogeny)  on  the  one  hand,  and  with  the  Lamarckian 
view  of  evolution  on  the  other,  led  him  to  some  inferences  which  have 
been  seriously  questioned.  Yet  some  of  these  inferences  proved  fertile,  use- 
ful, and  independent  of  the  grounds  they  were  built  on.  Haeckel's  bio- 
genetic  law  helped  Freud  in  elaborating  and  using  the  Jacksonian  model. 
Lamarck's  model  enabled  him  to  conceive  of  processes  of  adaptation 
for  which  Darwin's  theory  did  not  provide  the  conceptual  means.  The 
core  of  Freud's  genetic  conception,  namely  psychosexual  development,  is 
probably  the  most  familiar  segment  of  psychoanalytic  theory,  and  we 
need  not  dwell  on  the  evidence  which  led  Freud  to  make  it  the  center 
of  his  genetic  model.  The  model's  usefulness  was  not  exhausted  by  Freud : 
both  Hartmann's  [157]  concept  of  "change  of  function"  and  Erikson's 
[58,  62,  66]  "epigenetic"  conception  (which  extends  the  postulate  of 
"the  lawfulness  of  ontogeny"  to  behavior  development  far  beyond  the 
confines  of  psychosexual  development)  are  based  on  it. 

But  this  sketchy  statement  does  small  justice  to  the  pervasive 
significance  of  the  genetic  model  in  Freud's  theory  [see  Rapaport,  279]. 
Actually,  concepts  as  high  in  the  theoretical  hierarchy  as  identification 
and  transference,  and  theories  as  complex  as  that  of  object  choice  have 
their  roots  in  this  model.  It  was  the  genetic  model  which  enabled 
psychoanalysis — unlike  contemporary  learning  theories — to  put,  instead 
of  prior  learning,  prior  inborn  givens  in  the  center  of  its  conception  of 
learning.  [For  similar  attempts,  see  Lorenz,  225;  Tinbergen,  315;  Piaget, 
254;  Schiller,  303.]  Such  genetic  considerations  made  it  possible  for 
Freud  to  realize  the  significance  of  early  experiences  for  adult  behavior. 
It  took  academic  psychology  fifty  more  years  to  come  to  this  realization. 
[Cf.  Hunt's  confirming  experiment,  186,  187,  Hebb's  theory,  169,  and 
Beach  and  Jaynes'  review,  10]. 

Erikson's  [61]  as  well  as  Hartmamrs  [157]  and  his  collaborators' 
[167]  work  has  advanced  our  genetic  understanding,  as  has  Hartmann 
and  Kris's  discussion  [166]  of  the  genetic  and  the  dynamic  propositions 
of  psychoanalysis.  Werner's  [322]  and  Piaget's  [254,  255,  256]  work  in 
genetic  psychology  were  advances  in  the  same  direction.  Normative  and 
longitudinal  studies  have  contributed  considerable  systematic  observa- 
tional material  concerning  genetic  sequences.  Yet  the  methodological 
problems  involved  in  the  study  of  such  sequences  and  in  the  application 
of  the  genetic  point  of  view  have  still  not  been  solved.20 

20  See  John  Benjamin,  "Prediction  and  Psychopathological  Theory,"  in  press. 


d.  The  Jacksonian  (or  neural  integration  hierarchy]  model.  Accord- 
ing to  this  model,  the  nervous  system  consists  of  a  hierarchy  of  integra- 
tions in  which  the  higher  ones  inhibit  or  control  the  lower,  and  damage 
to  or  suppression  of  the  higher  ones  reinstates  the  function  of  the  lower. 
When  Freud  abandoned  his  neurological  anchorage  (1898),  he  ceased 
pursuing  neuropsychological  speculations  and  hypothesized  hierarchically 
organized  psychological  systems  patterned  on  Jackson's  hierarchy  of 
neural  levels  [98,  p.  488].  This  is  implied  in  one  of  the  specifications 
of  the  reflex-arc  model,  namely,  in  the  sequence  of  the  Systems  Un- 
conscious, Preconscious,  Conscious.  Freud's  Jacksonian  model  is  closely 
related  to  both  the  genetic  and  the  topographic  models,  and  its 
utility  is  that  it  provides  the  means  for  coordinating  systematically  those 
behavior  phenomena  which  are  not  attended  by  voluntary  control 
and/ or  consciousness  with  those  which  are.  Not  only  are  the  concepts  of 
the  Systems  Unconscious,  Preconscious,  and  Conscious  (as  well  as  those 
of  the  id,  ego,  and  superego)  organized  according  to  this  model,  but 
Freud  assumed  that  every  advancement  in  psychic  organization  goes 
along  with  a  new  censorship  [117,  pp.  122-127],  and  his  conception  of 
the  multiple  layering  of  defenses  within  the  ego  also  follows  the  same 
pattern  [116,  131]. 

But  this  does  not  exhaust  the  unique  significance  of  this  model  in 
Freud's  theory.  After  all,  Janet  and  Prince  based  their  conceptions  on 
a  similar  model,  even  if  Janet  did  not  assume  that  the  "subconscious" 
existed  under  the  control  of  consciousness  (in  the  Jacksonian  sense), 
but  rather  that  the  "subconscious"  was  created  by  dissociation  caused 
by  degeneration  and  precipitated  by  trauma.  In  Freud's  theory,  in- 
hibition of  lower  levels  by  higher  ones  served  as  the  model  for  the 
conceptualization  of  conflict.  Thus  inhibition  became  a  dynamic  event: 
the  result  of  a  clash  of  forces.  To  begin  with  (1895),  these  forces  were 
conceptualized  as  the  libidinal  affects  vs.  the  ego,  the  latter  being  the 
"ruling  ideational  mass"  which  serves  reality,  society,  and  morality  [35, 
p.  116].  Later  ( 1900),  this  conception  of  conflict  yielded  to  that  of  drives 
vs.  censorship,  the  latter  representing  ego  (self -preservative)  drives  [101, 
114].  The  final  conception  (1923)  was  that  of  the  interstructural  con- 
flict between  the  ego  and  the  id,  with  the  participation  of  the  superego 
on  one  or  both  sides  [126,  137].  Thus,  the  Jacksonian  model  coordinated 
those  observations  which  of  old  were  labeled  "conflict"  with  those  from 
which  "unnoticed  conflicts"  could  be  inferred,  and  it  served  as  the 
foundation  for  the  concepts  of  unconscious  conflict,  inhibition,  un- 
conscious drive  forces  and  counterforces,  which  led  to  the  theory  of 
symptoms,  and  ultimately  to  the  theory  of  mental  structure.  In 
summary:  the  Jacksonian  model  served  as  the  foundation  for  the 

The  Structure  of  Psychoanalytic  Theory  71 

dynamic  point  of  view  in  Freud's  theory,  and  also  contributed  to  the 
development  of  the  topographic  and  structural  points  of  view. 

e.  A  combined  model.  Let  us  now  sketch  in  more  detail  the  entropy 
(or  economic)  model,  the  behavior  sequence  restlessness  ->  sucking  on 
the  breast  -»  subsidence  of  restlessness.  [For  specific  references,  see  Rapa- 
port,  267.]  Here  restlessness  is  considered  the  referent  of  tension  accumu- 
lation, sucking  on  the  breast  that  of  tension-reducing  action  on  the  object, 
and  subsidence  of  restlessness  that  of  a  state  of  reduced  tension.  These 
are  equated  with  accumulation  of  cathexis,  action  on  the  cathected 
object,  and  discharge  of  cathexis,  which  in  turn  are  referred  to  drives 
reaching  threshold  intensity,  drive  action  on  the  drive-cathected  object, 
and  drive-gratification.  Finally,  the  direction  implicit  in  all  these  se- 
quences is  conceptualized  as  the  pleasure  principle. 

This  is  the  primary  model  of  action  (conation).  It  is  an  action 
model  because  it  does  not  account  for  thoughts  or  affects.  It  is  a  primary 
model  because  it  represents  only  actions  motivated  by  basic  drives,  with- 
out that  intervention  of  psychic  structures,  derivative  drives,  and  other 
motivations,  which  is  characteristic  of  most  observed  actions.  It  is  the 
first  of  the  six  models  to  be  derived  here  from  the  behavior  sequence 
which  is  considered  to  be  the  model  of  all  motivated  behavior.  We 
shall  now  derive  the  primary  models  of  thought  and  affect,  and  then 
turn  to  the  secondary  models. 

The  primary  model  of  cognition  (ideation)  was  formulated  by  Freud 
[98,  pp.  509-510,  533]  in  1900:  drive  reaching  threshold  intensity-^ 
absence  of  drive  object  ->  hallucinatory  idea  of  previous  gratification. 
When  the  drive  object  is  absent,  drive  action  is  not  possible,  and  a 
short  cut  to  hallucinatory  gratification  takes  place.  Drive  cathexis  is 
displaced  to  the  memory  of  past  gratifications,  bringing  these  to 
hallucinatory  intensity.  The  short  cut  and  direction  implicit  in  this  model 
were  conceptualized  by  Freud  as  wish-fulfillment.  It  is  worth  noting  that 
both  the  pleasure  principle  and  wish-fulfillment  (which  is  its  cognitive 
equivalent)  are  abstractions  remote  from  the  common-sense  meaning 
of  pleasure  and  wish.  The  model  extends  the  economic  point  of  view 
to  cognitive  phenomena,  and  its  concept  of  wish-fulfillment  expresses  the 
directed,  intentional  character  of  cognition.  This  model  makes  it  possible 
to  include  phenomena  like  dreams,  hallucinations,  illusions,  daydreams, 
reveries  in  the  theory  of  motivated  behavior,  and  serves  as  the  founda- 
tion for  those  concepts  which  in  the  secondary  model  of  cognition  co- 
ordinate these  thought  forms  with  the  more  familiar  cognitive  phe- 
nomena of  ordered  veridical  thought.  It  provides  the  theoretical  matrix 
for  the  understanding  of  free  associations  and  projective  techniques,  and 
concepts  for  the  explanation  of  the  observations  in  states  of  need  (hunger 


and  thirst  [197]),  stimulus  deprivation  [see  Hebb's  and  his  associates', 
25,  170,  171,  172,  and  Lilly's  experiments,  222,  223],  and  hypnotic 
states  [see  Gill  and  Brenman,  33,  148;  also  Rapaport,  280]. 

The  primary  model  of  affect  was  formulated  by  Freud  [98,  pp.  520- 
521]  in  1900:  drive  reaching  threshold  intensity-*  absence  of  drive 
object  -»  affect  discharge.  In  the  absence  of  the  drive  object,  drive 
action  not  being  possible,  emergency  discharge  through  affect-discharge 
channels  takes  place.  Freud  characterized  affect  discharges  first  as  "sally 
gates33  for  drive  tension  [98,  p.  520],  and  later  (1911)  as  discharges  into 
the  interior  of  the  organism  (autoplastic  adaptation),  in  contrast  to 
alterations  of  external  reality  by  action  (alloplastic  adaptation)  [108,  p. 
16].21  While  other  psychological  theories  postulate  direct  links  between 
affective  stimuli  and  the  bodily  changes  and  subjective  experiences  in- 
volved in  affects,  this  model — like  the  topographic  one — inserts  un- 
conscious ideas  and  drives  between  affective  stimuli  and  affective  re- 
sponses. This  modification  makes  possible  a  unified  theory  which  can 
account  for  anxiety  and  other  persistent  affects — for  affects  which  are  not 
triggered  by  obvious  affective  stimuli,  as  well  as  for  the  commonly  treated 
forms  of  affect.  It  also  eliminates  some  of  the  puzzles  confronting,  and 
various  of  the  contradictions  between,  the  familiar  theories  of  affect 
(James-Lange,  Cannon,  etc.).  [For  a  detailed  discussion,  see  Rapaport, 

The  relationship  between  the  primary  model  of  action  and  the 
primary  models  of  cognition  and  affect  is  indicated  by  the  presence  of 
the  drive  object  in  the  former  and  its  absence  in  the  latter.  The  rela- 
tionship of  the  cognition  model  to  the  affect  model  is  expressed  in  the 
combined  primary  model  of  cognition  and  affect  which  was  formulated 
by  Freud  [116,  p.  91;  117,  p.  Ill]  in  1915:  drive  at  threshold  inten- 
sity -»  absence  of  drive  object  — >  hallucinatory  idea  and /or  affect  dis- 
charge. It  was  devised  to  account  for  a  set  of  clinical  observations. 
Clinically,  the  repressed  drive  is  inferred  from  its  ideational  and  affect 
representations.  While  in  obsessional  ideas  only  the  ideational  repre- 
sentation of  the  drive  is  observed  (its  affect  representation  usually  suc- 
cumbing to  defense,  e.g.,  repression,  isolation,  or  displacement),  in 
hysterical  attacks  only  its  affect  representation  is  manifest  (the  ideational 
representation  succumbing  to  defense,  usually  repression).  Affect  and 
idea  are  thus  conceived  of  as  complementary  and/or  alternative  drive 

These  primary  models  unify  the  traditional  trichotomy  of  conation, 
cognition,  and  affection.  They  are  clearly  entropic  (economic)  models, 

21  Of  the  interactions  between  environment  and  organism,  those  which  result 
primarily  in  changes  of  the  organism  are  called  autoplastic,  and  those  which  result 
primarily  in  changes  of  the  environment  are  called  alloplastic. 

The  Structure  of  Psychoanalytic  Theory  73 

though  they  do  imply  the  topographic  (reflex-arc)  model  in  their  direc- 
tion, and  the  dynamic  model  in  the  role  played  by  the  drives.  They 
have  adaptive  implications  in  that  they  posit  coordinations,  guaranteed 
by  evolution,  between  the  drive  and  a  real  object  (the  drive  object), 
and  between  individuals  by  means  of  affect  expressions.  They  have 
structural  implications:  thresholds,  which  must  be  reached  by  drive 
intensities  before  drive  action  can  take  place,  and  drive-discharge 
channels.  They  have  genetic  implications,  in  that  they  are  assumed  to 
pertain  primarily  to  early  developmental  phases.  The  secondary  models 
elaborate  these  adaptive,  structural,  and  genetic  implications. 

The  secondary  model  of  action  was  outlined  by  Freud  [98,  pp.  533- 
534]  in  1900:  drive  reaching  threshold  intensity  ~->  derivative  drive 
mobilized  by  basic  drive  or  reaching  threshold  intensity  ->  structuralized 
delay  in  the  presence  of  the  drive  object  ->  detour-activity  searching  for, 
and  means-activity  reaching  fory  the  drive  object  ->  satisfaction.  Let  us 
take  the  relationships  between  the  steps  of  this  sequence  one  by  one. 

Drive  reaching  threshold  intensity  ->  derivative  drive  mobilized  by 
basic  drive  or  reaching  threshold  intensity.  This  step  implies  both  the 
Darwinian  and  the  Jacksonian  models:  it  has  both  genetic  and  hier- 
archic-structural implications.  In  the  course  of  development,  drives 
differentiate  into  a  hierarchy  of  derivative  drives.  This  drive  hierarchy, 
in  turn,  has  both  adaptive  and  structural  implications.  Derivative  drives 
differentiate  according  to  ontogenetic  laws,  e.g.,  that  of  psychosexual 
development,  yet  the  occasions  for  this  differentiation  are  environmental 
(for  instance,  the  periodic  unavailability  of  the  drive  object,  the  appear- 
ance of  substitute  objects,  the  environment's  response  to  and  demands 
for  new  ontogenetic  achievements,  etc.  [61,  62,  66])  and  the  differentia- 
tion itself  is  adaptive.  The  progressive  lessening  and  change  of  maternal 
and  familial  care  are  the  environmental  counterparts  of  this  adaptive 
development  and  provide  the  occasions  for  it.  Thus,  the  development  of 
derivative  drives  is  not  a  matter  of  pure  learning,  nor  is  it  blindly 
regulated  by  ontogenetic  laws. 

When  the  drive  object  is  absent,  the  drive-discharge  threshold  is 
raised  by  counter cathexes}  and  these  countercathectic  energy  distributions 
are  conceptualized  as  control  and  defense*2  structures  and  derivative 
drives.23  This  is  the  outstanding  dynamic  implication  of  the  model:  in- 

22  The  expression  "control  and  defense"  refers  to  an  insufficiently  studied  set 
of   phenomena.    Certain  countercathectic   energy  distributions   effectively  prevent 
the  execution  of  the  motivations  against  which  they  are  directed:  they  are  termed 
defenses.    Others    merely    delay,    modulate,    and    channel    motivations:    they    are 
termed  controls.  In  actual  observation,  instead  of  this  sharp  dichotomy,  we  find  a 
fluid  transition.  For  a  further  discussion  see  [268,  part  7]. 

23  The  manifestations  of  energy  distributions  are  always  forces,  and  clinically, 
defenses  are  always  recognized  by  the  appearance  of  new  motivations. 


hibition,  resulting  from  a  conflict  between  motivating  forces  and  struc- 
tures, gives  rise  to  new  motivating  forces. 

A  further  implication  of  this  first  relationship  is  that  action  may  be 
initiated  in  several  ways.  The  basic  drive  may  initiate  it  on  reaching 
threshold  intensity.  In  this  case  the  derivative  drives  (if  they  have  not 
reached  threshold  intensity)  may  either  be  bypassed,  or  triggered  by  the 
basic  drive.  But  action  may  also  be  initiated  directly  by  a  derivative 
drive  which  has  reached  threshold  intensity.  Or  it  may  be  initiated  when 
an  external  stimulation  provides  the  excess  excitation  which  brings  a 
basic  drive,  or  any  derivative  drive,  to  threshold  intensity.  Here  again 
we  encounter  the  hierarchic  arrangement,  and  the  short-circuiting  pos- 
sibilities discussed  in  connection  with  the  topographic  model. 

Now  to  the  second  relationship:  derivative  drive  mobilized  by  basic 
drive  or  reaching  threshold  intensity  -»  structuralized  delay.  This  rela- 
tionship implies  all  the  above-discussed  possibilities  for  initiating  action. 
Here  the  structuralized  delay  plays,  on  the  one  hand,  the  role  played 
by  the  drive-discharge  threshold  in  the  primary  model:  it  delays  dis- 
charge up  to  a  certain  point.  On  the  other  hand,  it  plays  the  same  role 
as  does  the  absence  of  the  drive  object  in  the  primary  model:  it  enforces 
delay  beyond  the  point  of  the  original  discharge  threshold.  Structuralized 
delay  (i.e.,  control  or  defense)  is  conceived  as  the  heightening  by  counter- 
cathexes  of  the  original  threshold,  so  that  the  object  of  the  drive  de- 
fended against  will  be  absent  (unnoticed  or  unusable)  from  the  point 
of  view  of  psychological  reality^  even  when  present  in  external  reality. 
Here  the  psychological  absence  of  the  object  plays  the  same  role  as  its 
real  absence  does  in  the  primary  model. 

Controls  and  defenses  are  conceptualized  as  structures:  their  rates 
of  change  are  slow  in  comparison  with  those  of  drive-energy  accumula- 
tions and  drive-discharge  processes.  The  delay  of  discharge,  which  these 
structures  make  possible,  is  the  crucial  distinction  between  the  primary 
and  the  secondary  models.  In  the  primary  models,  the  pleasure  principle 
(the  direct  discharge  tendency) 'prevails;  while  here,  the  contrary  princi- 
ple of  least  effort — which  is  one  of  the  referents  of  the  higher-order 
concept  reality  principle  (Freud,  1911) — prevails.  Threshold  and  drive 
intensity  are  relative  to  each  other,  yet  observations  necessitate  the  as- 
sumption that  the  control  and  defense  structures  may  become  relatively 
independent  of  the  drives.  The  relevant  observations — for  example,  the 
adaptive  role  of  some  behaviors  which  originated  as  defensive  reaction 
formations — are  the  same  as  those  upon  which  the  concept  of  autonomy 
rests.  The  structures  here  are  ego  structures;  their  autonomy  is  one  of 
the  implications  of  ego  autonomy  [157,  266,  280],  which  is  akin  to  All- 
port's  [8]  conception  of  "functional  autonomy." 

Now  to  the  next  relationship :  structuralized  delay  ->  detour-activity 

The  Structure  of  Psychoanalytic  Theory  75 

searching  for  and  means-activity  reaching  for  the  drive  object.  In  the 
primary  model,  when  the  drive  object  is  absent,  either  the  memory  of 
prior  gratification  is  drive-cathected,  resulting  in  hallucinatory  wish-ful- 
fillment, or  part  of  the  drive  cathexis  is  discharged  through  the  "sally 
gate"  of  affect-discharge  channels.  In  the  secondary  model,  stracturalized 
delay  postpones  discharge,  and  makes  possible  detours  from  the  direct 
route  of  gratification  and  search  for  the  drive  object.  The  concepts  of 
delay  and  detour  are  familiar  to  psychology  (Hunter,  1913;  Kohler, 
1917).  Though  Freud  introduced  them  in  1900,  psychologists  apparently 
did  not  notice  that,  once  taken  seriously  as  concepts,  they  can  account 
for  the  distinction  as  well  as  the  link  between  impulsive  and  controlled 
behavior.  We  will  encounter  delay  and  detour  again  in  the  secondary 
model  of  cognition. 

Finally,  the  last  relationship,  detour-  and  means-activity  -»  satis- 
faction, implies  that  it  is  not  necessarily  drive-gratification  that  is  at- 
tained by  sequences  of  this  sort.24  We  have  seen  that  such  sequences  may 
be  initiated  by  drives  or  derivative  drives,  either  of  which  may  be 
triggered  by  external  excitations.  Now — taking  the  autonomy  of  the  ego 
into  consideration — we  must  add  that  external  excitation  may  also  di- 
rectly trigger  detour-  and  means-behaviors:  the  functions  subserving 
detour-  and  means-behavior  are  ego  functions.  Structuralized  delay  and 
detour,  and  structure  in  general  (defense-,  control-,  and  means-struc- 
tures) are  the  concepts  which  enable  this  theory  to  account  for  tension 
maintenance  and  tension  increase,  and  not,  as  is  generally  supposed  [7], 
for  tension  reduction  only.  The  shift  from  the  "gratification"  of  the  pri- 
mary model  to  the  "satisfaction"  of  this  model  indicates  that  full  dis- 
charge of  drive  tension  gives  way  to  discharge  compatible  with  the 
maintenance  of  tension  which  is  made  inevitable  by  structure  formation. 

The  secondary  model  of  cognition  was  outlined  by  Freud  [98,  pp. 
509-510,  533-536]  in  1900:  drive  or  derivative  drive  at  threshold  in- 
tensity ->  Structuralized  delay  -» experiment  in  thought  with  small 
cathectic  amounts  to  anticipate  and  plan,  locate  and  act  upon  the  drive 
object.  The  first  two  steps  in  this  model  are  the  same  as  those  in  the 
secondary  model  of  action,  and  the  considerations  presented  above  apply. 
The  relationship  between  Structuralized  delay  and  experimental  action  in 
thought  is  the  only  one  to  be  discussed.  According  to  the  primary  model 
of  cognition — which  follows  the  pleasure  principle — when  drive  action 
cannot  take  place,  a  short  cut  to  hallucinatory  gratification  occurs, 
through  the  mechanisms  of  displacement,  condensation,  substitution, 

24  K.  Lewin's  [216,  220,  221]  quasi  needs  are  examples  of  this.  More  generally, 
the  distinction  between  drive-gratification  and  satisfaction  corresponds  to  the  dis- 
tinction between  the  varieties  of  action-initiation  discussed  above  on  p.  74.  For 
further  discussion  see  [267]. 


symbolization,  etc.  In  the  secondary  model,  under  the  same  conditions, 
structuralized  delay  prevents  the  short  cut,  and  detour-behavior  involv- 
ing ordered  thinking  results. 

The  conception  of  two  types  of  memory  organization  rests  on  this 
difference  between  the  two  models  [268,  note,  pp.  630-631].  In  the 
drive  organization  of  memories,  all  the  memorial  (ideational)  representa- 
tions of  a  drive  are  organized  around  it  and  are  equivalent  to  each 
other.  The  syncretic  mechanisms  enumerated  above  express  this  equiv- 
alence.25 In  the  conceptual  organization  of  memories  the  equivalences 
have  two  determiners:  on  the  one  hand,  empirical  coordinations  (fre- 
quent contiguity),  on  the  other,  logical  implications  (not  all  frequent 
contiguities  are  admitted,  but  only  those  which  are  compatible  with 
logical  implications).26 

These  two  memory  organizations  do  not  predicate  two  classes  of 
thought,  but  conceptualize  two  different  aspects  of  any  given  thought, 
with  the  stipulation  that  the  conceptual  organization  is  hierarchically 
higher  than  the  drive  organization,  and  has  a  controlling  function  over 
it.  These  cognitive  models,  and  their  genetic  relation  to  each  other, 
represent  the  first  consistent  attempt  to  coordinate,  within  one  theory, 
those  forms  of  thought  (obsessions,  delusions,  dreams,  etc.)  which  are 
peremptory  and  those  (practical  thought,  rational  thought,  rigorously 
logical  thought)  which  we  can  take  or  leave. 

In  this  secondary  model,  the  intentional,  anticipatory  potential  of 
thought  derives  from  the  directedness  of  the  primary  model,  while  its 
realistic  efficacy  derives  from  the  structuralized  delay  which  militates 
against  the  immediate  discharge  and  gratification  tendency  of  the  pri- 
mary model,  and  thus  permits  the  development  and  use  of  conceptual 

The  secondary  model  of  affect  was  formulated  by  Freud  [131,  chap. 
8,  particularly  pp.  76-79]  in  1926:  drive  or  derivative  drive  at  threshold 
intensity  -»  structuralized  delay  -»  affect  signal  released  by  the  ego  from 
structurally  segregated  affect  charges.  [A  more  detailed  discussion  will  be 
found  in  "The  Psychoanalytic  Theory  of  Affects,"  274,  and  in  Organi- 
zation and  Pathology  of  Thought,  268.]  The  first  relationship  of  this 
sequence  is  identical  with  that  of  the  other  secondary  models,  and  the 
considerations  advanced  above  apply,  with  one  exception :  here  the  role 

25  The  term  "equivalence"  as  used  here  is  a  generalization  of  the  equivalence 
implied  in  the  concept  of  "equivalent  stimuli."  It  pertains  not  only  to  stimuli  but 
to  responses  also.  It  applies  not  only  to  "nondiscriminable"  stimuli  or  responses 
but  also  to  those  whose  relationship  to  each  other  is  that  of  indicator  to  indicated. 

25  Further  discussion  of  these  memory  organizations  will  be  found  in  Organiza- 
tion and  Pathology  of  Thought  [268]  and  in  "The  Psychoanalytic  Theory  of 
Thinking"  [265], 

The  Structure  of  Psychoanalytic  Theory  77 

of  the   drive's  reaching  threshold  Intensity  is  changed  and  the  con- 
sequences of  mounting  drive  tension  are  anticipated  by  the  ego. 

Early  in  his  theory-building,  Freud  assumed  that  when  no  drive 
action  can  take  place,  the  affect-discharge  channels  serve  as  "sally  gates53 
for  part  of  the  dammed-up  drive  cathexes  (affect  charge).  As  late  as 
1915  he  assumed  [1 17,  pp.  109-1 12]  that  only  the  drive  intensity  and  the 
capacity  of  the  affect-discharge  channels  determine  the  affect  charge, 
and  that  the  latter,  before  its  discharge,  is  not  segregated  from  the  drive 
cathexes.  This  conception  still  applies  to  affects  in  early  phases  of 

Later  (1926),  however,  it  became  necessary  to  assume  that  a  struc- 
tural segregation  of  affect  charges  from  drive  cathexes  at  large  takes 
place,  parallel  with  the  development  of  the  motivational  and  structural 
hierarchy,  as  specific  affects  and  affect-discharge  channels  differentiate  at 
each  level  of  the  hierarchy.  Originally,  accretion  of  drive  intensity  was 
assumed  to  use  the  affect-discharge  channels  automatically  when  drive 
action,  in  the  absence  of  the  drive  object,  is  not  possible;  now,  the 
absence  of  objects  has  been  internalized  in  the  form  of  structuralized 
delay,  and  the  ego  structures  subserving  this  delay  include  such  as 
keep  the  affect  charge  segregated  and  control  its  discharge  also.  The 
segregated  affect  charges  are  therefore  under  the  control  of  the  ego: 
when  rising  drive  tension  impinges  on  the  ego's  defense  structures,  the 
ego  uses  the  segregated  affect  charge  to  give  an  anticipatory  affect  signal, 
which — though  of  small  intensity  in  comparison  to  affect  discharge — 
mobilizes  (by  virtue  of  the  pleasure  principle)  countercathexes  to  rein- 
force the  defenses,  and  thus  prevents  drive  discharge  [131,  pp.  18-20, 
112-117].  Affects  change  in  the  course  of  ontogeny  from  discharge 
phenomena  into  signals,  from  safety  valves  for  drive  tension  into  antic- 
ipations of  the  means  for  preventing  drive  discharge.  Under  such 
"normal"  circumstances  as  bereavement  or  danger  (but  also  when  ex- 
posed to  wit  and  drama),  as  well  as  under  pathological  conditions,  the 
signal  affects  may  yield  their  place  to  discharge  affects  [see  75].  Also, 
the  segregated  affect  charge,  like  all  cathectic  amounts,  may  manifest 
itself  as  a  motivating  force.27  According  to  the  secondary  model,  affects 
may  serve  as  discharge  processes,  as  anticipatory  ego-signals  for  mount- 
ing drive  tension,  and  as  motivations;  thus  it  unites  a  wide  variety  of 
observations  concerning  emotions.  Indeed,  most  of  the  observations  on 

ST  For  instance,  actions  related  to  an  unconscious  sense  of  guilt  may  be  motivated 
by  the  aggressive  impulse  which  gave  rise  to  the  guilt  affect,  or  by  a  (derivative) 
motivation  which  arose  as  a  reaction-formation  to  this  aggressive  impulse,  or  by 
the  guilt  affect  itself  which  has  attained  the  status  of  a  relatively  autonomous 
motivation  [see  267].  This  may  be  a  link  to  Leeper's  [211]  motivation  theory  of 


affects  which  the  various  academic  theories  account  for — or  fail  to  ac- 
count for — are  coordinated  in  this  model. 

The  behavior  forms  represented  by  the  secondary  models  arise 
according  to  ontogenetic  laws  from  those  represented  in  the  correspond- 
ing primary  models,  but  their  development  also  depends  on  the  en- 
vironmental conditions  and  is  thus  adaptive. 

In  contrast  to  the  primary  models.,  all  the  secondary  models  involve 
structuralized  delay,  that  is  to  say,  progressive,  hierarchically  layered 
structure  development.  The  structures  in  question  are:  defense  and 
control  structures,  structures  which  segregate  affect  charges,  and  the 
means  structures  which  subserve  secondary  action-  and  thought-processes. 
A  parallel  development  takes  place  in  the  hierarchy  of  motivations :  each 
step  in  structure  development  results  in  a  delay  imposed  on  motivations., 
which  in  turn  gives  rise  to  new  derivative  motivations  and  affects.  This 
multifaceted  hierarchic  development  is  the  development  of  the  ego 
[268,  276]  and  involves  the  differentiation  of  the  ego  from  the  id,  and 
the  superego  from  the  ego.  The  id-ego-superego  trichotomy  is  the 
broadest  structural  articulation  of  the  mental  organization  and,  as  such, 
a  crucial  conception  of  the  clinical  theory  of  psychoanalysis.  Since  it 
can  be  derived  from  the  models  discussed,  it  is  not  an  independent  model 
and  we  shall  not  dwell  on  it  here.  [For  a  similar  conception,  see  Glover, 

The  secondary  models  lean  heavily  on  the  Darwinian  (genetic)  and 
the  Jacksonian  (hierarchic)  models,  and  therefore  the  structural,  genetic, 
and  adaptive  considerations  are  central  to  them.  Yet  they  also  include 
the  topographic,  economic,  and  dynamic  considerations  of  the  primary 
models.  Thus,  this  combined  model,  which  is  an  elaboration  of  the 
entropic  (economic)  model,  does  unite  all  the  models  Freud  used.  But  it 
does  so  at  the  price  of  falling  into  six  partial  models  which  by  their  very 
nature  (if  not  by  that  of  the  theory,  or  even  of  the  subject  matter) 

4.  The  comprehensiveness  of  empirical  reference.  From  the  very  be- 
ginning, the  theory  implied  a  comprehensive  empirical  reference, 
though  it  centered  on  the  psychology  of  drives  and  primary  processes, 
and  maintained  that  its  findings  concerning  these  were  of  unrestricted 
validity.28  Freud  asserted  as  late  as  1917  [121,  pp.  330-333]  that  the 
postponement  of  the  exploration  of  secondary  processes,  ego  functions, 
reality  relations,  and  adaptation  was  a  deliberate  policy  and  not  a  failure 

28  That  is  to  say,  they  are  the  ultimate  determiners  of  all  behavior.  The 
conception  of  ultimate  determiner,  and  the  restrictions  imposed  on  it  later,  will 
be  discussed  further  on  pp  93  fT.,  below. 

The  Structure  of  Psychoanalytic  Theory  79 

to  recognize  their  importance.  Actually,  Freud  did  take  steps  toward  in- 
cluding them  in  the  theory  in  1900  [98,'  pp.  533,  535]  and  in  1911  [108], 
and  consolidated  these  steps  between  1917  [121]  and  1926  [131].  But 
only  in  1939  did  Hartmann  [157]  give  the  first  systematic  formulation  of 
reality  relationships  and  adaptation  by  expanding  the  frame  of  reference 
of  ego  psychology.  This  systemization,  which  was  accompanied  by  the 
claim  that  psychoanalysis  is  a  comprehensive  system  of  psychology,  was 
continued  in  the  studies  of  Hartmann,  Kris,  Loewenstein  [160,  161,  162, 
167,  168,  206],  Rapaport  [268,  especially  part  7;  267,  277],  Jacobson 
[190,  191,  192],  and  Gill  and  Rapaport  [149]. 

In  the  meanwhile,  and  even  before  these  developments  in  the  main- 
stream of  psychoanalysis,  adaptation  and  reality  relationships,  especially 
the  role  of  interpersonal  relations  and  society,  were  central  to  the  theories 
of  Adler,  Horney,  Sullivan,  and  Kardiner  [see  Munroe,  240].  Erikson 
[56,  57,  58,  59,  60,"  61,  62,  63,  66]  was  the  first  to  unite  this  tributary  of 
theoretical  development,  which  enlarged  the  actual  realm  of  empirical 
reference,  with  the  mainstream  of  the  theory.29 

Finally,  in  the  late  thirties,  forties,  and  fifties,  the  influence  of  psy- 
choanalysis and  of  the  new  psychoanalytic  ego  psychology  expanded  to 
the  whole  of  psychology,  first  through  projective  techniques  into  clinical 
psychology,  then  into  experimental  clinical  psychology,  and  finally  into 
experimental  psychology  proper  [see  309].  Thus  the  original  claim  of 
comprehensiveness  for  this  theory  is  gradually  being  realized. 

If  we  must  single  out  an  outstanding  limitation  of  this  theory's  claim 
to  comprehensiveness,  then  we  should  choose  its  lack  of  a  specific  learning 
theory.  Psychoanalysis  has  created  grounds  on  which  contemporary 
learning  theories  (Hull,  Bollard,  Miller,  Mowrer,  etc.)  can  be  sharply 
criticized,  and  its  conception  of  the  primary  process  (e.g.,  the  drive 
organization  of  memories)  and  of  the  secondary  process  (e.g.,  the  con- 
ceptual organization  of  memories)  can  be  regarded  as  foundations  for  a 
theory  of  learning.  But  like  Lewinian  and  Gestalt  psychology,  it  has  failed 
to  offer  a  specific  alternative  learning  theory.  Though  Hartmann's  [157] 
automatization  concept  seems  to  open  a  new  approach  to  the  problem 
of  learning — as  did  K.  Lewin's  [221]  ossification  concept — so  far  no  one 
has  used  it.  The  problem  of  learning — how  a  process  turns  into  a 
structure,  or  in  other  words,  the  long-term  survival  and  availability  of 
experience — has  not  been  solved  by  psychoanalysis  either. 

5.  Quantification  and  mensuration.  Psychoanalytic  theory  does  con- 
tain quantitative  considerations  (particularly  in  its  economic  point  of 
view),  but  the  translation  of  these  into  actual  measurements  presents 

29  Hartmann,  whose  work  is  an  indispensable  link  between  Erikson's  work  and 
classic  psychoanalytic  metapsychology,  laid  the  metapsychological  groundwork 
for  this  unification  but  did  not  actually  undertake  it. 


difficulties  which  have  not  been  overcome.  Some  of  these  difficulties  will 
be  mentioned  here  and  in  Section  V.  below. 

Freud  did  not  take  a  theoretical  stand  implying  unquantifiability;30 
yet  neither  he  nor  any  other  psychoanalyst  tried  to  quantify  the  variables 
of  the  theory.  However  useful  and  indispensable  the  theory  is  clinically, 
however  much  light  it  sheds  on  a  broad  range  of  human  phenomena,  and 
however  consistently  everyday  and  clinical  experience  confirms  its  help- 
fulness, as  a  theory  it  requires  exact  tests  of  confirmation  which  in  turn 
require  the  mathematization  of  the  relationships  posited  by  it.  The 
obstacles  to  mathematization  are:  (a)  The  basic  independent  variable 
(drive  cathexis  in  general,  libido  in  particular)  postulated  by  the  psycho- 
analytic theory  is  an  intrapsychic  one,  related  to  organic  changes  and 
intrapsychic  structural  conditions,  rather  than  to  external  stimuli;  thus 
it  is  hard  to  manipulate31  and  measure.  ( b )  The  avenues  through  which 
such  variables  may  exert  their  causal  effect  are  multiple  and  interchange- 
able (cf.  Tolman's  vicarious  function,  Lewin's  substitute  tasks,  and 
Heider's  equifinality),  and  thus  hard  to  predict,  observe,  and  measure.32 
(c)  The  distance  between  the  theory's  major  variables  and  the  observed 
phenomena  makes  it  uncertain  whether  or  not  any  measure  obtained 
actually  quantifies  a  particular  variable. 

But  these  obstacles  need  not  prevent  mathematization  (e.g.,  quanti- 

30  We  do  know,  however,  that  he  occasionally  took  a  practical  stand  to  the 
effect  that  the  theory  needs  no  experimental  confirmation  [cf.  Rosenzweig,  293]. 
I  am  indebted  to  Dr.  Saul  Rosenzweig  for  a  personal  communication  which  indi- 
cates that  Freud,  in  a  conversation  with  H.  A.  Murray  and  in  one  with  R.  Grinker, 
seems  to  have  retracted  this  stand. 

31  Not  the  least  of  the  obstacles  to  manipulation  is  the  inviolable  privacy  of  the 

32  The  point  is  frequently  made  that  Freud's  failure  to  quantify  his  variables 
was  due  to  his  having  come  from  "another  tradition,"  and  that  the  continued 
avoidance  of  quantification  stems  from  the  development  of  psychoanalysis  "apart 
from  academic  psychology."  True,  Freud's  neurological  research  was  in  the  area 
of  nonquantitative  neuroanatomy.  It  is  also  possible  that,  as  Holt  [177]  suggests, 
Freud's    experience    with    Fliess's    "numbers    game" — combined    with    a    general 
limitation  in  mathematical  thinking,  which  he  mentions  repeatedly  [94] — made 
him  averse  to  quantitative  considerations.  Still,  these  arguments  seem  to  miss  the 
mark,  and  distract  attention  from  the  lack  of  quantitative  methods  applicable  to 
intrapsychic  variables.  Academic  psychology  has  only  recently  begun  to  be  con- 
cerned with  such  methods.  In  addition,  these  arguments  disregard  that  the  Helm- 
holtz  tradition  was  the  matrix  of  both  Freudian  and  academic  psychology,  and  that 
the  biology  of  that  time  was  not  centered  on  quantification  but  rather  on  the 
significance  of  the  single  case  and  on  the  tracing  of  genetic  connections.  The  theory 
of  evolution  seems  to  have  been  a  "good  science"  even  though  the  complex  statistics 
applied  to  it  by  G.  G.  Simpson  [311,  312]  were  not  available  to  Darwin.  In  just 
what  sense  Darwin's  and  Freud's  theories  are  good  sciences  is  an  interesting  but  so 
far — to  my  knowledge — unanswered  question. 

The  Structure  of  Psychoanalytic  Theory  81 

fication),  though  they  do  suggest  that  the  road  to  It  will  be  long  and 
arduous.  It  is  doubtful  that  the  long  hierarchic  chain  of  intermediary  con- 
cepts interposed  between  the  major  explanatory  constructs  and  the  ob- 
servables  can  be  bypassed,  and  that  direct  relations  can  be  found  between 
them.  This  highlights  the  importance  of  theory  construction,  since  only 
a  tightly  built  theory  (with  clearly  stated  definitions  and  implicative 
rules)  can  support  confirming  tests  on  observables  which  are  at  a  great 
distance  from  the  constructs:  the  models  discussed  above  show  that  in 
this  theory  the  implicative  rules  change  with  the  distance  from  the  basic 

Many  features  of  observables  can  be  counted,  rated,  and  measured, 
but  the  observables  alone  cannot  tell  us  which  features  and  what  method 
of  counting  or  measuring  them  will  reveal  the  relationship  between  them 
and  the  explanatory  constructs:  only  theory  can  do  that.  A  certain 
amount  of  trial-and-error  (ad  hoc]  quantification  is  inevitable,  but  it 
will  never  yield  a  theory:  theory  is  the  product  of  theory-making.  The 
confirmation  or  refutation  of  a  theory  requires  that  we  quantify  those 
features  of  the  observables  which  correspond  to  the  dimensions  of  the 
theory's  variables.  The  ad  hoc  quantifications  are  not  necessarily  useless; 
they  may  be  the  means  by  which  the  deductions  from  the  theory  and  the 
inductions  from  the  observables  are  brought  progressively  closer  to  each 
other,  and  by  which  the  essential  measurables  are  progressively  selected 
from  the  multitude  of  all  measurables.  But  this  selection  cannot  be 
achieved  by  blind  measurement  unguided  by  theory:  there  is  no  end  to 

The  first  steps  toward  quantification  are  (a)  systematic  mastery  of 
the  theory  as  it  exists  at  present,  (b]  systematic  attempts  to  tighten  the 
theory,  (c]  the  selection  of  measurables  relevant  to  the  variables  of  the 
theory.  So  far,  no  attempt  at  quantification  has  included  these  steps. 
Most  of  the  experimenters  who  have  attempted  to  confirm  or  refute  the 
relationships  posited  by  psychoanalytic  theory  were  unaware  of  the 
nature  of,  and  the  variables  involved  in,  the  relationships  which  they  set 
out  to  test. 

This  may  seem  to  be  a  sad  picture  of  the  theory  and  a  summary 
indictment  of  the  experimenters  who  have  tried  to  deal  with  it.  Neither 
of  these  is  intended.  We  are  blinded  by  the  rapid  development  of  new 
sciences  in  our  time.  The  rapid  growth  of  biochemistry  and  biophysics 
was  possible  because  they  had  the  solid  foundations  of  several  thousand 
years  of  physics  and  chemistry.  Some  psychologists  are  bent  on  linking 
psychology  to  those  sciences  now,  hoping  for  an  equally  spectacular 
growth  of  psychology.  Others  are  more  patient.  They  do  not  deplore  the 
present  state  of  the  theory,  nor  consider  the  experimenters  to  be  fools. 

33  For  similar  considerations  in  geology,  see  Rich  [287]. 


In  their  eyes  these  difficulties  are  phenomena  of  a  very  early  phase  in  the 
development  of  a  science.  Clinical  observation  shows  that  conscious  in- 
formation does  not  eliminate  symptoms  rooted  in  unconscious  forces  and 
that  conscious  intent  is  usually  no  substitute  for  the  lack  of  unconscious 
motivation.  Likewise,  we  may  assume  that  consciously  borrowed  method- 
ological sophistication,  however  much  it  may  help  otherwise  in  develop- 
ing psychology,  cannot  circumvent  the  long  and  time-consuming  process 
all  sciences  have  gone  through.  The  process  of  development  which  brings 
about  the  interplay  between  the  observables  and  the  theories  is  always 
slow.34  Quantification  and  methodological  sophistication  are  late  prod- 
ucts of  any  science  and  as  such  they  should  be  long-range  goals:  mis- 
taking them  for  proximal  goals  can  render  a  science  impotent. 

6.  Formal  organization.  The  expositions  of  psychoanalytic  theory 
have  been  informal  rather  than  systematic;  in  the  main  they  were  di- 
rected by  internal  consistency  within  the  theory  and  between  observables 
and  the  theory.  In  the  last  twenty  years  attempts  at  systematic  formula- 
tion [26,  73,  150,  157,  166,  167,  267,  268,  274]  have  been  made,  but  no 
hypothetico-deductive  system-building  is  in  sight. 


In  order  to  discuss  the  systematic  independent,  intervening,  and 
dependent  variables  of  the  psychoanalytic  theory,  it  seems  necessary  to 
sketch  the  theory's  structure. 

A.  The  Subject  Matter  of  Psychoanalysis  Is  Behavior  (the  Empiri- 
cal Point  of  View) 

This  proposition  has  often  been  overlooked,  probably  because  the 
theory's  stress  on  unconscious  processes  and  drives,  psychological  struc- 
tures, dynamics  and  economics  obscured  the  fact  that  it  conceives  of  all 
of  these  as  explanatory  concepts  of  behavior. 

34  We  have  some  idea  why  this  process  is  so  slow.  If  logic,  methodology,  and 
mathematics  were  the  pacemakers  of  development  in  sciences,  this  development 
could  be  fast  enough  in  psychology.  But  the  pacemaker  is  not  methodology — it  is 
human  invention.  ("Developmental  projects,"  "crash  programs,"  and  "inter- 
disciplinary teams"  are  effective  only  in  highly  developed  sciences  or  else  in 
situations  where  the  makeshifts  of  pooled  ignorance  are  the  most  that  can  be 
had.)  Methodology,  since  it  deals  with  relationships  of  concepts,  all  of  which  are 
potentially  valid,  can  go  on  continuously,  building  ever-new  "castles  in  Spain."  But 
human  invention  consists  of  discontinuous  events,  each  of  which  requires  long 
preparation,  since  in  it  an  individual's  thought  patterns  must  come  to  grips  with 
patterns  of  nature,  and  only  those  rare  encounters  in  which  a  unique  human 
thought  pattern  actually  matches  a  unique  pattern  of  nature  will  matter.  If  the 
match  is  not  specific  and  precise,  or  if  the  individual  is  not  prepared  to  recognize 
it,  or  if  he  does  recognize  it  but  is  not  ready  to  use  it,  the  moment  is  lost. 

The  Structure  of  Psychoanalytic  Theory  83 

Behavior  in  this  theory  is  broadly  defined,  and  includes  feeling  and 
thought  as  well  as  overt  behavior,  "normal"  as  well  as  "pathological/3 
frequent  as  well  as  unique  forms  of  behavior.  This  corollary  too  has 
often  been  overlooked,  probably  because  of  the  stress  in  psychoanalytic 
literature  on  "latent  behavior"  and  on  pathology,  both  of  which  served 
as  points  of  departure  for  the  theory.  Indeed,  not  before  Hartmann's 
[157]  major  study  (1939)  was  it  directly  stated  that  psychoanalysis  is  a 
general  psychology  which  embraces  the  study  of  normal  as  well  as 
pathological  behavior,35  though  the  principle  of  the  thoroughgoing  psy- 
chological determination  of  all  behavior  has  been  the  cornerstone  of  the 
psychoanalytic  theory  from  the  beginning  and  was  explicitly  stated  in 
1905  [99]. 

Thus,  all  appearances  to  the  contrary  notwithstanding,  psychoanalysis 
does  not  differ  from  other  theories  in  its  view  of  the  subject  matter  of 
psychology  (though  it  defines  behavior  far  more  comprehensively  than 
most),  nor  in  its  assumption  of  determinism  (though  it  probably  de- 
manded this  earlier  and  in  a  more  sweeping  fashion).  Yet  it  does  differ 
from  other  psychologies  in  assuming  psychological  determinism,  and  in 
its  stress  on  "latent  behavior"  in  general  and  on  the  unconscious  deter- 
minants of  behavior  in  particular  (cf.  Section  II.  E.,  below). 

B.  Behavior  Is  Integrated  and  Indivisible:  The  Concepts  Con- 
structed for  Its  Explanation  Pertain  to  Different  Components  of  Be- 
havior and  Not  to  Different  Behaviors  (the  Gestalt  Point  of  View) 

In  the  clinical  parlance  (and  even  in  the  theoretical  writings)  of  psy- 
choanalysis, the  explanatory  concepts  are  anthropomorphized,  reified,  or 
at  best  presented  in  existential  terms,  giving  the  impression  that  they 
refer  to  entities  or  at  least  that  each  of  them  refers  to  a  specific  behavior. 
But  this  is  not  consistent  with  the  theory.  The  tendency  to  anthropo- 
morphize and  reify,  and  the  preference  for  hypothetical  constructs  prob- 
ably derives  from  clinical  practice,  where  there  is  a  premium  on  the 
"plausibility"  and  "uncomplicated  everyday  application"  of  concepts. 

In  concrete  terms:  no  behavior  can  be  described  as  an  id  behavior, 
or  an  ego  behavior,  or  a  conscious  behavior.  These  concepts  all  refer  to 
specific  aspects  of  behaviors  and  not  to  specific  behaviors.  Every  behavior 
has  conscious,  unconscious,  ego,  id,  superego,  reality,  etc.,  components. 
In  other  words,  all  behavior  is  multiply  determined  (overdetermination). 
Since  behavior  is  always  multifaceted  (and  even  the  apparent  absence 
of  certain  facets  of  it  requires  explanation),  the  conception  of  multiple 

35  It  is  noteworthy,  though,  that  Freud's  recently  discovered  manuscript  [94, 
appendix],  which  is  the  predecessor  of  the  theory  of  psychoanalysis,  has  the  scope 
of  a  general  psychology. 


determination  (or  overdetermination)  may  be  regarded  as  a  purely 
formal  consequence  of  this  method  of  conceptualization.  This  naturally 
does  not  rule  out  the  possibility  that  the  conception  of  overdetermination 
is  also  required  by  the  nature  of  the  observations;  in  fact  overdetermina- 
tion as  a  concept  was  originally  introduced  [35,  pp.  156,  219]  in  refer- 
ence to  observations,36  rather  than  on  purely  theoretical  grounds.  From 
the  very  beginnings  of  psychoanalytic  theory,  observations  made  the  con- 
cept of  overdetermination  both  necessary  and  central.  Academic  psy- 
chologies did  not  develop  such  a  concept,  probably  because  their  methods 
of  investigation  tend  to  exclude  rather  than  to  reveal  multiple  determina- 
tion. But  they  did  not  escape  the  problem  itself:  every  behavior  phe- 
nomenon has  perceptual,  learning  (memorial),  conceptual  (cognitive), 
motor,  etc.,  components;  and  the  rival  psychological  theories  (perceptual 
theory  of  cognition,  learning  theory  of  perception,  motor  theory  of 
thought,  etc. )  show  both  the  presence  of  the  problem  and  the  confusion 
resulting  from  a  failure  to  face  it  squarely.37 

36  For  example,  when  a  subject  executes  the  posthypnotic  suggestion  to  shut  a 
door  and  explains  that  he  did  so  because  of  the  draft  [53],  then  his  action  is  de- 
termined both  by  the  hypnotic  suggestion  of  which  he  is  not  conscious  and  by 
his  conscious  intention  to  escape  the  draft.  Dr.  A.  B.  Wheelis   (San  Francisco) 
suggests    (personal  communication)    that  there  are  distinctions  among  overdeter- 
mination,  multiple   determination,   and  multiple  levels  of  analysis,  which  hinge 
on  whether  the  determiners  are  independent  and  sufficient  causes  of  the  behavior 
in   question    (overdetermination)    or   not    (multiple   determination).    It   appears, 
however,  that  in  psychoanalytic  theory  neither  such  independence  nor  such  suffi- 
ciency of  causes  can  be  demonstrated  or  perhaps  even  defined.  The  fact  that  to 
escape  the  draft  would  be,  under  other  conditions,  a  sufficient  cause  for  shutting 
the  door  does  not  make  it  a  sufficient  cause  in  the  posthypnotic  situation.  The 
matter  of  the  "independence"  of  causes  is  an  autonomy  issue  (cf.  Section  II.  H., 
below).   Overdetermination,  to  my  mind,  implies  precisely  such  a  lack  of  inde- 
pendence and  sufficiency  of  causes  and  is  inseparably  connected  with  the  multiple 
levels  of  analysis  necessitated  by  this  state  of  affairs.  Mr.  J.  Zsoldos  (Maabaroth, 
Israel)    suggests    (personal   communication)    that   the    "overdetermination"    issue 
crops  up  where  "weak  (sensitive)   systems'5  are  exposed  to  overwhelmingly  large 
forces,  that  under  such  conditions  simple  functional  relationships  do  not  obtain, 
and  quantitative  analysis  is  possible  only  in  terms  of  statistics;  so  that  "weak  sys- 
tems" have  only  statistics,  not  "laws."  This  suggestion  seems  to  imply  that  the 
overdetermination  issue  is  the  psychological  counterpart  of  the  controversy  between 
Einstein's  theory  and  present-day  atomic  physics.  The  psychoanalytic  theory  of 
overdetermination  as  it  stands — if  I  read  it  correctly — implies  laws  and  not  statis- 
tics. To  use  Einstein's  phrase,  "The  good  Lord  does  not  play  dice"  in  this  theory 
either.  Nevertheless,  the  possibility  of  a  statistical  interpretation  of  overdetermina- 
tion must  be  kept  open,  even  if  reluctantly;  a  specific  and  workable  statistical 
interpretation  would  be  preferable  to  an  interpretation  which  assumes  the  existence 
of  laws  but  does  not  specify  any  implicative  rules  and  thus  permits  neither  con- 
firmation nor  refutation. 

37  "Field  theories"  may  be  looked  upon  as  attempts  to  meet  this  problem. 

The  Structure  of  Psychoanalytic  Theory  85 

This  psychoanalytic  proposition  has  implications  over  and  above 
that  indivisibility  of  behavior  from  which  the  considerations  pertaining 
to  overdetermination  stem.  It  requires  that  each  of  the  conceptually 
differentiated  aspects  of  behavior,  as  well  as  the  spatial  and  temporal 
context  of  behavior,  be  treated  as  an  integrated  whole.  But  we  need  not 
pursue  this  point  further:  it  seems  to  coincide  grossly  with  the  general 
postulate  of  Gestalt  psychology. 

C.  No  Behavior  Stands  in  Isolation:  All  Behavior  Is  That  of  the 
Integral  and  Indivisible  Personality  (the  Organismic  Point  of  View) 

This  thesis  demands  that  the  explanation  of  any  behavior  fit  into  the 
theory  of  the  workings  of  the  total  personality.38  Freud's  most  direct 
statement  of  this  thesis  is  probably  that  pertaining  to  dreams.  Once  he 
had  developed  the  theory  of  dreams,  he  raised  the  question:  what  kind 
of  theory  of  personality  could  embody  this  dream  theory.  In  Chapter  7 
of  The  Interpretation  of  Dreams  he  proceeded  to  construct  the  frame- 
work of  such  a  theory  of  personality  [98,  pp.  469,  470,  485-486]. 

Yet  this  implication  of  psychoanalytic  theory,  too,  has  been  overlooked 
by  many  psychoanalysts  and  psychologists,  probably  because  the  stress 
on  the  central  role  of  drives  made  it  appear  to  the  psychoanalyst  that  the 
fundamental  drives  sufficiently  guarantee  the  unity  of  behavior  and  per- 
sonality, and  gave  the  psychologist  the  impression  that  in  this  theory  the 
"atomistically"  conceived  behavior  fragments  are  held  together  only  by 
the  "glue"  of  the  drive  concept.  The  organizing,  integrative  role  of  the 
secondary  process  (1900),  however,  speaks  eloquently  against  both  of 
these  views  [98,  pp.  533-536],  and  the  "structural  point  of  view"  (ego, 
id,  superego,  etc.),  which  clearly  embodies  principles  of  cohesiveness 
other  than  drives  [126,  pp.  15-18;  see  also  Nunberg,  245,  and  Erikson, 
63],  should  have  dispelled  these  misconceptions.  It  did  not. 

To  be  sure,  it  is  easy  to  find  passages  in  Freud  which  lend  them- 
selves to  atomistic  interpretation,  but  the  theory  itself  does  not.  French 
[87,  88,  89]  devoted  his  major  work  to  demonstrating  the  role  of  the 
integrative  field  in  psychoanalytic  considerations.  Psychoanalytic  studies 
in  psychosomatics  embraced  Goldstein's  organismic  view  as  "consistent" 

38  It  may  be  objected  that  Freud  did  not  explicitly  formulate  the  organismic 
point  of  view  and  that  only  the  organismic  biologists  and  Wertheimer,  Goldstein, 
and  Wheeler  arrived  at  it.  But  our  task  here  is  not  limited  to  a  collation  of  Freud's 
explicit  systematic  formulations.  Recognition  of  the  organismic  thesis  of  psycho- 
analysis is  the  more  important,  since  Gestalt  as  well  as  personalistic  psychologists 
viewed  psychoanalysis  as  an  atomistic  and  mechanistic  theory.  Wertheimer  was 
vehement  about  this  in  his  lectures  and  conversation,  and  G.  Allport  [7]  outspoken 
in  his  writings.  The  attitude  of  many  practicing  psychoanalysts  in  regard  to  sym- 
bols, dream  interpretations,  etc.,  to  the  contrary  notwithstanding,  this  thesis  ap- 
pears to  be  a  basic  implication  of  Freud's  theory. 


with  psychoanalysis.  The  projective  techniques,  which  developed  under 
the  impact  of  psychoanalytic  conceptions,  borrowed  from  psychoanalysis 
the  projective  postulate  [259,  261]  that  all  behavior  is  integral  to,  and 
characteristic  of,  the  behaving  personality.  Nevertheless,  this  implica- 
tion of  psychoanalysis  remained  so  remote  that  French  [85]  in  1933 
(before  his  familiarity  with  K.  Lewin),  and  later  Mowrer  [239],  as  well 
as  Bollard  and  Miller  [48],  found  it  feasible  to  link  psychoanalysis  to  the 
atomistic  conditioned-response  theory  of  learning. 

What  this  organismic  point  of  view  asserts  is  not  that  each  behavior 
is  a  microcosm  which  reflects  the  macrocosm  of  the  personality,  but 
rather  that  an  explanation  of  behavior,  in  order  to  have  any  claim  to 
completeness,  must  specify  its  place  within  the  functional  and  structural 
framework  of  the  total  personality  and,  therefore,  must  include  state- 
ments about  the  degree  and  kind  of  involvement,  in  the  behavior  in 
question,  of  all  the  relevant  conceptualized  aspects  of  personality. 

D.  All  Behavior  Is  Part  of  a  Genetic  Series,  and  through  Its  Ante- 
cedents, Part  of  the  Temporal  Sequences  Which  Brought  About  the 
Present  Form  of  the  Personality  (the  Genetic  Point  of  View) 

This  thesis  implies  that  every  behavior  is  an  epigenetic  product  [58] 
and  thus  can  and  must  be  studied  genetically  for  its  full  explanation 
[166].  However,  it  implies  neither  a  view  of  behavior  as  the  "matura- 
tion53 of  a  preformed  behavior  repertory,  nor  one  according  to  which  be- 
haviors "develop"  from  accumulating  experience;  rather,  it  views  be- 
havior as  the  product  of  an  epigenetic  course  which  is  regulated  both  by 
inherent  laws  of  the  organism  and  by  cumulative  experience. 

The  genetic  point  of  view  does  not  conflict  with  K.  Lewin's  in- 
sistence that  only  forces  and  conditions  which  are  here  and  now  present 
can  in  the  here-and-now  exert  an  effect;39  it  asserts  simply  that  much  of 
what  "exists"  here  and  now  in  the  subject  can  only  be  known  through  a 
genetic  exploration  of  its  antecedents.  This  implies  that  descriptively 
identical  behaviors  may  differ  in  their  psychological  significance,  depend- 
ing on  their  genetic  roots.  But  it  also  implies  that  the  empirical  relevance 
of  a  behavior  to  a  situation  in  which  it  occurs  alone  does  not  necessarily 
explain  it  and  that  the  explanation  must  also  take  into  consideration 
the  epigenetic  laws  which  brought  the  behavior  about.  Indeed,  it  is 
peculiar  that  it  should  have  been  Lewin  who  criticized  the  genetic  point 
of  view,  when  he  more  than  any  other  psychologist  stressed  the  distinc- 
tion between  genotype  and  phenotype  and  sharply  criticized  the  use  of 
achievement  concepts.  He  gave  the  example :  identical  typewriting  speeds 

39  Nor  does  it  clash  with  Lewin's  [218]  and  Chein's  [43]  point  that  the  past 
reconstructed  by  the  patient  in  psychoanalysis  is  the  past  as  he  views  it  in  the 

The  Structure  of  Psychoanalytic  Theory  87 

of  applicants  for  a  job  provide  insufficient  information,  since  they  may 
be  products  of  maximal  exertion  or  routine  approach,  disuse  or  peak 
efficiency,  recent  training  or  established  working  level  [220,  pp.  89-91]. 
We  must  conclude  that  without  the  exploration  of  its  genetic  antecedents, 
a  behavior  can  only  be  described  in  terms  of  achievement  concepts. 

The  genetic  point  of  view  refers  to  the  history  of  the  drive  proc- 
esses which  express  themselves  in  a  given  behavior,  to  the  history  of 
the  structures  (e.g.,  those  corresponding  to  "abilities")  used  in  it,  and  to 
the  history  of  the  subject's  relation  to  the  situation  in  which  the  be- 
havior occurs.  An  example  of  the  distinctions  implied  here:  a  sudden 
attack  of  stammering,  which  is  brought  about  by  defense  against  an 
aggressive  impulse.  A  genetic  exploration  will  take  into  consideration 
those  past  experiences,  and  the  controlling  structures  crystallized  from 
them,  which  modulated  the  development  of  the  aggressive  drive  and  thus 
gave  the  power  to  arouse  aggression  to  situations  like  the  one  which 
aroused  the  subject's  anger  in  this  instance.  It  will  also  consider  the  past 
experiences  which  led  to  defense  against  aggression  in  general  or  against 
that  particular  kind  of  aggression  which  came  into  play  in  the  given 
situation.  It  will  extend  to  those  past  experiences  which  made  the  verbal 
avenue  for  the  expression  of  anger  particularly  vulnerable  to  defense  and 
to  those  past  events  which  shaped  the  stammer,  that  is,  the  form  which 
defense  took  in  this  situation.  In  each  of  these  instances,  experience 
denotes  both  the  historical  event  in  its  external  setting  and  the  internal 
situation  of  the  subject,  including  the  specific  phase  of  his  maturation 
and  development. 

Although  the  genetic  point  of  view  does  not  refer  specifically  to  the 
contextual  (spatial-temporal)  determination  of  behavior,  it  does  imply 
contextual  determination.  Moreover,  it  does  specifically  refer  to  the  intra- 
psychic  context :  to  the  contemporary  state  of  the  personality  as  a  whole 
and  as  a  genetic  product. 

The  genetic  character  of  the  psychoanalytic  theory  is  ubiquitous  in 
its  literature.  The  concept  of  "complementary  series"40  is  probably  the 
clearest  expression  of  it:  each  behavior  is  part  of  a  historical  sequence 
shaped  both  by  epigenetic  laws  and  experience  [101,  summary;  279]; 
each  step  in  this  sequence  contributed  to  the  shaping  of  the  behavior 
and  has  dynamic,  economic,  structural,  and  contextual-adaptive  relation- 
ships to  it.  Such  complementary  series  do  not  constitute  an  "infinite 
regress":  they  lead  back  to  a  historical  situation  in  which  a  particular 
solution  of  a  drive  demand  was  first  achieved,  or  a  particular  apparatus 
was  first  put  to  a  certain  kind  of  use  [cf.  166]. 

But  this  formulation  is  incomplete  because  it  disregards  those  ob- 
servations to  which  the  concept  of  autonomy  refers  [157,  161,  280].  Cer- 

*  Erganzungsreike,  see  Freud  [121,  lecture  21]  and  Fenichel  [73,  pp.  121ff.]. 


tain  behaviors  do  (all  tend  to)  cease  to  be  shaped  further  by  their  re- 
currence: they  become  automatized  [157]  and  relatively  autonomous 
from  their  genetic  roots;  they  take  on  a  tool-  or  means-character  and  at- 
tain a  high  degree  of  stability.  However,  automatization  and  autonomy 
make  not  only  for  stability,  but  also  for  the  availability  of  the  behavior  as 
a  means  of  adaptive  performance.  These  automatized  behaviors  can  also 
be  studied  genetically,  but  from  that  point  at  which  they  become  auto- 
matized, their  "complementary  series"  proves  relatively  unrevealing, 
since  from  there  on  situation  and  context  may  bring  them  into  action, 
even  in  the  absence  of  the  motivations  which  gave  rise  to  them  originally. 
In  Section  II.  H.,  pp.  93-97,  we  return  to  the  concepts  of  automatization 
and  autonomy.  They  are  akin  to  Woodworth's  [329,  pp.  lOOff.]  concept 
of  "habits  as  drives,"  and  to  Allport's  [6;  8,  pp.  76ff.]  concept  of  func- 
tional autonomy,  but  they  are  more  specific  and  more  differentiated  than 
either  of  these. 

Psychoanalysis  as  a  genetic  psychology  deals  with  the  genetic  roots 
of  behaviors,  with  the  degree  of  autonomy  behaviors  attain,  and  with  the 
genetic  roots  of  the  subject's  relation  to  the  reality  conditions  which 
codetermine  the  appearance  of  a  behavior  at  a  given  point  in  the  person's 
life.  Yet  the  first  formal  statement  of  the  genetic  point  of  view  of  psy- 
choanalysis is  that  of  Gill  and  Rapaport  [149]. 

E.  The  Crucial  Determinants  of  Behaviors  Are  Unconscious  (the 
Topographic  Point  of  View) 

Per  se,  this  thesis  is  not  alien  to  any  psychology,  and  particularly  not 
to  those  psychologies  which  exclude  all  phenomena  of  consciousness  from 
their  subject  matter,  and  thus  have  to  assume  that  the  determinants  of 
behavior  are  extraconscious.  All  psychologies  deal  with  conditions  "un- 
noticed" by  the  subject,  and  with  "unnoticed"  or  "unnoticeable"  proc- 
esses underlying  his  behavior.  The  psychoanalytic  thesis  of  unconscious 
determination,  however,  differs  from  these  [98,  pp.  543-544;  110]  in 
several  respects:  (1)  it  explicitly  conceptualizes  that  which  is  unnoticed 
or  unnoticeable  [110];  (2)  it  asserts  that  the  unnoticed  or  unnoticeable 
can  be  inferred  from  that  which  is  noticed  by  the  subject  (and/or  the 
observer),  by  means  of  the  effects  of  the  unnoticed  and/or  the  unnotice- 
able upon  that  which  is  noticed  [99] ;  (3)  it  asserts  that  the  rules  govern- 
ing the  noticed  are  different  from  those  governing  the  unnoticed,  and 
that  the  unnoticed  can  be  inferred  by  considering  the  deviations  of  the 
noticed  from  its  usual  patterns  [117,  pp.  118-122];  (4)  it  makes  a 
systematic  distinction  between  the  unnoticed  and  the  unnoticeable  (the 
unnoticed  can  become  conscious,  whereas  the  unnoticeable,  by  definition, 
cannot) ;  it  expresses  this  distinction  by  the  terms  "descriptive"  vs, 
"dynamic"  unconscious,  and  conceptualizes  it  as  the  distinction  between 

The  Structure  of  Psychoanalytic  Theory  89 

the  Systems  Pr^conscious  and  t/rcconsclous  [98,  pp.  543-544];  (5)  while 
other  psychologies  treat  the  unnoticeable  in  nonpsychological  terms  (brain 
fields,  neural  connections,  etc.),  psychoanalysis  consistently  treats  it  in  the 
psychological  terms  of  motivations,  affects,  thoughts,  etc. 

The  differences  between  the  laws  governing  the  Conscious  and  those 
governing  the  Unconscious  are  expressed  in  the  concept  of  primary  and 
secondary  processes.41  In  the  early  phases  of  the  theory,  the  Cs  and  the 
Ucs  were  considered  systems  of  paramount  significance.  Later  on  ( 1923) 
they  were  subordinated  to  the  structural  conception  id-ego-superego 
[126],  and  still  later  (1938)  were  relegated  to  the  role  of  "qualities" 
[140].42  In  keeping  with  Freud's  early  formulations  [35,  98],  recent 
contributions  to  ego  psychology  treat  consciousness  as  a  superordinate 
sense  organ.  They  attribute  to  it  a  complex  hierarchic  layering  (states 
of  consciousness)  [268,  270,  276],  and  thus  treat  it  on  a  level  of  abstrac- 
tion different  from  that  accorded  the  unconscious. 

F.  All  Behavior  Is  Ultimately  Drive  Determined  (the  Dynamic 
Point  of  View) 

This  thesis  of  psychoanalysis  has  become  only  too  well  known  in 
a  doubly  distorted  form :  all  behavior  is  determined  by  sex.  The  qualifier 
"ultimately"  was  omitted,  and  sex,  libido,  drive,  and  psychosexuality 
were  equated.  It  is  certainly  true  that,  until  recently,  the  drives  most 
closely  studied  by  psychoanalysis  were  the  sexual  drive  and  its  partial 
drives.  But  psychosexuality  was  defined  in  such  a  broad  way  that  it  was 
by  no  means  synonymous  with  "sex"  [101].  Self -preservative  and  ego  in- 
stincts were  also  discussed  early  [101,  114,  115,  121],  but  were  dropped 
later  since  they  did  not  prove  helpful  in  organizing  empirical  evidence. 
The  history  of  the  theory  of  drives  (narcissism,  instinctual  vicissitudes, 
life  and  death  instincts,  monistic  drive  theory,  aggressive  drive)  suggests 
that  the  early  centering  on  libidinal  drives  helped  Freud  to  explore  the 
nature  of  drives  and  their  motivational  role  [115],  but  did  not  settle  the 
theory  of  drives  itself  [26].  In  spite  of  some  recent  advances  [160,  168], 
it  is  still  unclear  how  many  and  what  kinds  of  drives  need  to  be 

The  crucial  role  attributed  to  libidinal  drives  is  not  a  theoretical 
necessity  in  this  system.  It  seems  to  derive  from  two  of  Freud's  major 
achievements :  the  conception  of  the  determination  of  behavior  by  drives 
and  the  observation  of  infantile  sexuality.  The  fact  that  the  theory  linked 
these  two  to  each  other  very  early  may  have  retarded  a  full  assessment 
of  the  role  of  libidinal  drives  in  psychological  life. 

41  The  relationship  of  Conscious  vs.  Unconscious  to  primary  vs.  secondary  proc- 
esses is  not,  however,  a  one-to-one  coordination. 

42  Section  I.  B.  3.  a.  presented  the  issues  discussed  here  from  another  angle. 


Infantile  sexuality  was  an  empirical  discovery  made  by  the  method 
of  psychoanalytic  reconstruction  and  verified  by  the  method  of  direct  ob- 
servation on  infants  and  children.  Though  of  empirical  origin,  it  grew 
to  systematic  importance  because  it  provided  the  example  for  the 
development  of  drives  [101]. 

The  broader  principle  of  drive  determination  was  an  empirical  as 
well  as  conceptual  discovery.  The  empirical  discovery  embodies,  besides 
its  novel  aspect,  two  familar  observations:  (1)  behavior  is  not  always 
triggered  by  external  stimulation  but  often  occurs  without  it,  as  though 
spontaneously;  (2)  behavior  (which  by  scientific  fiat  is  causally  deter- 
mined) evinces  a  goal-directedness,  a  purposive,  teleological  character. 
The  conceptual  discovery,  which  took  the  form  of  the  definition  of  the 
drive  concept,  was  the  first  large-scale  attempt  to  cope  with  both 
of  these  observations  simultaneously.  The  drive  is  defined  as  a  causal 
agent  inherent  in  the  organism  [115,  p.  64],  and  thus  it  can  account 
for  the  apparent  "spontaneity"  of  behavior.  Moreover,  since  the  definition 
makes  the  effectiveness  of  the  drive  dependent  on  an  environmental 
condition,  namely,  the  presence  of  the  drive  object,  it  can  also  account 
for  the  purposiveness  of  behavior.  This  coordination  of  drive  and  drive 
object — which  is  assumed  to  be  guaranteed  by  evolution — at  first 
tolerates  little  if  any  means-activity  and  demands  immediate  consum- 
mation (pleasure  principle).  In  the  course  of  development,  it  becomes 
more  flexible,  and  permits  delay  and  interpolation  of  means-activities, 
though  it  selects  and  organizes  these  in  the  service  of  consummation. 
Later  on  it  permits  substitute  goals  and  a  variety  of  means-  and  con- 
summatory-activities,  until  finally  it  prescribes  only  the  consummatory 
behavior,  and  provides  no  more  than  the  motivational  framework  for  in- 
strumental behavior.  This  conception  of  motivation  accounts  not  only 
for  the  spontaneity  and  teleology  of  behavior,  but  also  for  behavior 
elicited  by  external  stimulations,  since  the  latter  may  be  conceived  of  as 
drive  objects,  or  substitutes  for  them. 

If  psychological  theories  can  be  divided  into  two  classes  according 
to  whether  they  consider  the  human  psyche  a  tabula  rasa  on  which  ex- 
perience writes,  or  an  organization  of  actualities  and  potentialities  which 
limits  and  regulates  the  extent  and  kind  of  changes  that  experience  can 
bring  about,  then  the  drive  conception  certainly  belongs  to  the  latter 
class.  In  philosophical  systems  this  distinction  is  crudely  paralleled  by 
Hume  vs.  Kant  [cf.  9,  pp.  7ff.;  252,  254,  262].  In  terms  of  psychology 
the  distinction  is  paralleled,  for  instance,  by  the  conception  of  passive 
registration  of  experience  vs.  active  organization  of  it,  a  distinction  which 
involves  the  nurture-nature  controversy.  Psychoanalysis  was  one  of  the 
first  theories  to  recognize  the  interaction  of  nature  and  nurture  in  the 
development  of  behavior.  Drives  represent  the  "nature"  factor;  and 

The  Structure  of  Psychoanalytic  Theory  91 

their  vicissitudes,  in  the  course  of  experience,  the  interweaving  of  nature 
and  nurture.  Moreover,  the  coordination  of  drive  and  drive  object  ex- 
presses a  primary  coordination  given  by  evolution  between  human  nature 
and  its  environment  and  is  thus  a  psychological  representation  of  the 
biological  adaptedness  of  the  species  to  its  environmental,  ecological 

Finally  we  come  to  the  "ultimate  determination  of  behavior  by 
drives."  Here  we  meet  relationships  like  those  connected  with  the 
"genetic  point  of  view."  While  early  psychoanalysis  actually  maintained, 
without  reservation,  the  thesis  of  "ultimate  drive  determination,"  the 
increasing  evidence  for  the  "indivisibility  of  behavior"  led  to  the  realiza- 
tion that  behavior,  in  so  far  as  it  can  be  said  to  be  determined  by  drives, 
must  also  be  said  to  be  determined  by  defenses  and/or  controls.  More- 
over, with  the  development  of  ego  psychology,  the  question  was  no  longer 
which  of  these  was  the  ultimate  determiner  of  behavior  but  rather  in 
what  respect  and  to  what  extent  was  each  the  determiner  of  a  given  be- 
havior [cf.  Waelder,  320].  Finally,  behaviors  were  encountered  in  which 
drive  determination  was  in  abeyance.  This  led  to  the  concept  of  ego 
autonomy  (cf.  Section  II.  H.). 

Thus  the  thesis  of  the  ultimate  drive  determination  of  behavior, 
while  it  remains  valid  in  psychoanalysis,  must  be  regarded  in  the  con- 
text of  the  other  theses  here  discussed,  which  qualify  it  and  limit  its 
scope.  The  concepts  of  drive  fusion,  drive  differentiation  into  partial 
drives,  conflict,  etc.,  all  pertain  to  the  dynamic  point  of  view  and  in- 
dicate limitations  to  the  conception  of  ultimate  drive  determination. 

G.  All  Behavior  Disposes  of  and  Is  Regulated  by  Psychological 
Energy  (the  Economic  Point  of  View) 

This  thesis,  too,  has  a  history.  In  the  first  phase  of  psychoanalytic 
theory  (abreaction  theory — up  to  1898),  psychological  energy  was 
equated  with  affects,  and  the  "defenses"  which  prevented  abreaction  were 
not  conceptualized  in  economic  terms  [35,  98].  In  the  second  phase 
(1900-1926),  psychological  energy  was  conceptualized  as  drive  energy, 
and  the  methods  used  in  discharging  it  as  the  primary  process.  It  was 
recognized  that  other  (secondary)  processes,  using  minute  quantities  of 
energy,  exert  a  regulative  function  over  those  which  dispose  of  drive 
energies  [98,  particularly  chap.  7;  108,  116,  117,  119,  120].  The 
relationship  between  these  two  kinds  of  processes  was  conceived  much 
like  that  described  nowadays  as  obtaining  between  power  engineering 
and  information  engineering  [cf.  Wiener,  326,  pp.  53-56;  Rapaport, 
264],  In  this  phase,  however,  little  attention  was  paid  to  the  nature  and 
origin  of  the  secondary  process.  In  the  third  phase  (after  1926),  some 


understanding  was  gained  of  the  secondary  process,  of  its  congenital  roots, 
and  of  the  progressive  ontogenetic  transition  to  it  from  the  primary 
process  [131,  chap.  8,  especially  pp.  82,  112-117;  cf.  also  274]. 

All  behaviors  have  both  primary  process  and  secondary  process 
aspects,  though  one  or  the  other  may  predominate.  The  primary  process 
operates  with  drive  energies,  and  its  regulative  principle  is  the  tendency 
toward  tension  reduction  (pleasure  principle)  :43  it  strives  toward  imme- 
diate discharge  of  energy  accumulations,  by  a  direct  route  and  by  means 
of  the  mechanisms  of  displacement,  condensation,  substitute  formation, 
symbolization.  The  secondary  process  operates  by  the  principle  of  least 
action,  is  oriented  toward  objective  reality,  and  finds,  through  delays 
and  detours,  by  experimental  action  in  thought,  the  safest  course  toward 
the  sought-for  object  in  reality,  suspending  the  discharge  of  drive  energies 
until  the  object  has  been  found  [98,  pp.  533-536;  108]. 

In  the  course  of  development,  hierarchically  layered  structures  arise 
(defenses  and  controls)  which  act  as  "dikes."  These  not  only  delay  or 
prevent  discharge,  but  also  dimmish  the  drives5  tendency  toward  im- 
mediate discharge.  These  structures  are  conceived  of  as  built  by  "bind- 
ing" drive  energies  to  heighten  the  originally  given  drive-discharge 
thresholds  [98,  pp.  533-534;  116,  117].  Their  effect  of  diminishing  the 
drives'  tendency  toward  immediate  discharge  is  conceptualized  as 
"neutralization,"  special  instances  of  which  are  referred  to  as  delibidiniza- 
tion,  deaggressivization,  or  sublimation  [126,  pp.  61-65;  206,  164]. 
These  processes  of  binding  and  neutralization  make  cathexes  (hyper- 
cathexes,  attention-cathexes)  available  to  the  secondary  process,  to  be 
used  in  small  quantities  for  experimental  action  in  thought  [108,  p.  16], 
as  signals  in  the  form  of  affects  [98,  p.  536;  and  131],  and  as  counter- 
cathexes  (against  drives)  for  building  new  and  for  reinforcing  existing 
defensive  structures.  Once  a  process  of  "neutralization"  is  assumed,  the 
original  dichotomy  of  primary  and  secondary  processes  yields  to  a 
hierarchic  model  in  which  these  two  represent  theoretical  extremes  and 
the  actually  observed  phenomena  represent  intermediary  forms  [206; 
268,  e.g.,  p.  536].  The  energies  of  lesser  degrees  of  neutralization  (drive 
derivatives)  show  characteristics  of  their  drive  origin^  whereas  those  of 
higher  degrees  do  not,  and  are  at  the  disposal  of  the  ego.44  Sources  of 

43  We  will  not  discuss  here  the  distinction  between  an  "optimal35  and  a  "maxi- 
mal" lowering  of  tension.  The  latter  has  been  assigned  to  the  so-called  "nirvana 
principle"   (associated  with  the  "death  instinct") — a  speculative  excursion  which 
does  not  seem  to  be  an  integral  part  of  the  theory.  Cf.  Freud  [123],  Fenichel  [74], 
and  Hartmann,  Kris,  and  Loewenstein  [168]. 

44  While  the  ego  builds  new  or  employs  existing  defenses  against  drive  deriva- 
tives  of  low  neutralization,   it  can  make  use  of  derivatives  of  high  neutraliza- 
tion, since  it  can  deal  with  these  by  means  of  its  controls.  This  difference  seems  to 
be  akin  to  the  difference  between  all-or-none  vs.  graduated  processes. 

The  Structure  of  Psychoanalytic  Theory  93 

neutralized  energy  other  than  drives  are  also  compatible  with  the  theory 
[161,  pp.  86-87;  162,  p.  21]. 

These  psychological  energies  are  not  equated  with  any  known  kind 
of  biochemical  energy.  They  do  not  correspond  to  the  muscular  energy 
expended  in  overt  behavior.  The  differences  in  the  quality  (mobility  vs. 
degree  of  neutralization)  of  psychological  energy45  correspond  to  the 
observed  differences  between  overvalent  thoughts  (e.g.,  obsessions)  and 
passing  thoughts  (e.g.,  logical  thinking),  between  impulsive  or  com- 
pulsory actions  and  actions  which  are  a  matter  of  choice.  There  is  an 
obvious  conceptual  similarity  between  Freud's  energies  and  Lewin's 
tensions  and  between  Freud's  drives  and  drive  objects  and  Lewin's  forces 
and  object  valences,  though  there  are  also  significant  differences  be- 
tween them.  They  are  alike  in  that  they  cannot  be  expressed  in  the 
mathematical  formulas  in  which  physics  expresses  its  energy  concepts, 
yet  they  are  referents  of  phenomena  which  seem  to  abide  by  the  laws 
of  energy  exchanges — conservation,  entropy,46  least  action.  (However, 
it  is  neither  implied  nor  ruled  out  that  biochemical  energy  exchanges 
may  eventually  be  discovered  which  correspond  to  the  exchanges  of 
psychological  energy  inferred  from  behavior  by  psychoanalysis.)  They 
differ,  among  other  things,  in  that  Lewin's  concepts  do  not  account  for 
the  differences  in  quality  here  discussed. 

From  the  point  of  view  of  the  energy  economy  of  the  organism,  the 
exchanges  of  psychological  energy  may  be  considered  as  the  work  of  an 
information  engineering  network  which  controls  the  biochemical  energy 
output  of  overt  behavior.  But  this  network  itself  is  multiply  layered,  so 
that  ever  smaller  quantities  of  energy  control  the  networks  which  carry 
and  dispose  of  greater  quantites  of  energy  [cf.  Wiener,  326].  For  further 
discussions  pertinent  to  the  economic  point  of  view,  see  Sections  I.  B. 
3.  b.  andV.A. 

H.  All  Behavior  Has  Structural  Determiners  (the  Structural  Point 

of  View) 

The  simplest  way  to  put  the  issue  of  structure  is  to  point  out 
that  drive  energies  can  be  conceived  of  only  within  well-defined  systems 
which  have  definite  thresholds  of  discharge.  The  structural  conception 
could  well  have  been  first  necessitated  by  the  observations  pertaining 
to  such  discharge  thresholds;  and  the  prototype  of  the  conflict  between 

45  Speaking  about  the  "quality"  of  energies  does  not  contradict  the  fact  that 
energy  is  a  quantitative  construct.  Physics  too  speaks  of  different  kinds  of  energies: 
heat,  light,  etc.  But  in  psychology  we  do  not  yet  have  transformation  equations 
to  express  the  quantitative  relationships  of  these  qualities  to  each  other. 

46  The  validity  of  the  economic  point  of  view  is  unaffected  by  Bertalanffy's 
conception  of  "open  systems"  [20]. 


drives  and  structures  could  have  been  the  relationship  between  drives 
and  thresholds.  Actually.,  however,  this  was  not  the  origin  of  the  struc- 
tural point  of  view,  even  though  the  early  (1895)  neuropsychological 
form  of  the  theory  would  allow  such  an  interpretation  [94,  appendix], 
Recently  (1939)  Hartmann  [157;  see  also  162]  pointed  out  that  not  only 
thresholds  but  ego  apparatuses  like  memory,  perception,  and  motility  are 
also  structural  givens.47  But  the  structural  point  of  view  did  not  originate 
in  these  structural  givens  either. 

It  was  observed  that  drives  do  not  unequivocally  determine  behavior 
in  general,  nor  symptom  formation  in  particular.  In  contrast  to  the  drive 
processes,  whose  rate  of  change  is  fast  and  whose  course  is  paroxysmal, 
the  factors  which  conflict  with  them  and  codetermine  behavior  appeared 
invariant,  or  at  least  of  a  slower  rate  of  change.  The  observation  of 
these  relatively  abiding  determiners  of  behavior  and  symptom  seems  to 
have  been  the  foundation  on  which  the  concept  of  structure  was  built. 

In  the  first  phase  of  the  theory  (up  to  1900),  reality  was  considered 
the  factor  which  interfered  with  the  drives,  through  the  ego  (con- 
sciousness) in  general  and  through  its  defenses  in  particular  [96,  97]. 
But  this  view  did  not  reach  a  conceptual  status  at  the  time  and  was 
superseded  (1900)  by  the  conception  of  intrapsychic  censorship  exerted 
by  ego  (self-preservative)  instincts.  A  conception  of  psychological  life  as 
a  continuous  clash  of  drive  forces  arose  [98],  and  the  abiding  character 
of  the  interfering  factors  was  lost  sight  of.  Not  even  the  link  established 
between  the  concepts  of  censorship  and  the  secondary  process  [98,  108, 
117]  conceptualized  the  abiding  character  of  these  two  drive-controlling 
factors.  Instead,  Freud  again  (1911-1917)  became  concerned  with  the 
role  of  reality,  and  considered  it  to  be  the  factor  which  interferes  with 
the  drives  and  becomes  a  codeterminer  of  behavior.  Yet  he  still  assigned 
the  reality-testing  functions  to  the  secondary  process  and  to  the  ego 
drives  [108,  121].  But  the  further  study  of  censorship  (particularly  of 
its  repressive  function)  and  of  the  secondary  process  shed  additional  light 
on  these  interfering  factors:  they  were  now  conceptualized  as  counter- 
cathexes,  which  delay  the  discharge  of  drive  cathexes,  and  by  their 
permanent  deployment,  prevent  the  return  of  the  repressed  [116,  117]. 
This  formulation  of  a  permanent  deployment  of  countercathexes  is  the 
beginning  of  the  structural  conception. 

An  explicit  formulation  of  the  structural  conception  became  neces- 
sary when  it  was  realized  that  not  only  the  drives  but  also  most  of  these 
invariant  factors  which  interfere  with  drives  are  unconscious  [126].  The 

47  The  structural  givens  in  question  are  not  the  muscular  apparatuses  of  motility, 
nor  the  end  organs  of  perception,  etc.,  but  rather  their  psychological  regulations: 
for  instance,  those  psychological  structures  through  which  the  control  and  trigger- 
ing of  the  motor  apparatus  is  effected. 

The  Structure  of  Psychoanalytic  Theory  95 

topographic  conception,  which  relegated  all  crucial  determiners  (drives) 
of  behavior  to  the  Unconscious  and  all  epiphenomena  and  "apparent 
determiners"  (ego)  to  Consciousness,  became  systematically  untenable, 
and  was  replaced  by  the  structural  conception.  The  topographic  division 
into  the  Systems  Unconscious,  Preconscious,  and  Conscious  yielded  to 
the  structural  conception  of  id,  ego,  and  superego.  The  "ultimate  drive 
determiners"  were  conceptualized  as  id,  the  codetenniners  (whether 
conscious  or  not)  as  ego,  and  a  specialized  segment  of  the  ego  as 
superego.48  Whereas  the  id  was  conceived  of  as  the  congeries  of  drives 
(coexistent  even  if  contradictory),  the  ego  was  defined  as  a  cohesive 
organization,  whose  function  was  to  synthesize  the  demands  of  id, 
superego,  and  reality  [126].  The  ego  was  conceived  of  as  a  structure 
which  codetermines  (along  with  the  drives)  every  behavior,  and  is  re- 
sponsible for  the  coordinated  and  organized  character  of  all  behavior,  in- 
cluding specific  drive-discharge  actions  (e.g.,  sexual  intercourse).  But 
the  ego  as  a  structure  proved  so  complex  that  its  exploration  is  even  now 
only  just  beginning.  The  recognition  of  the  structure-building  and 
structural  role  of  identifications  [126]  was  followed  by  the  recognition 
of  the  role  of  the  ego's  defensive  substructures  [131].  In  addition  to  these 
two  kinds  of  substructures,  psychoanalytic  ego  psychology  also  came  to 
recognize  orienting  (perceptual),  processing  (conceptual),  and  executive 
(motor)  substructures,  when  it  was  realized  that  they  are  ready  tools 
(means)  available  to  ego  processes  [157,  266].49 

To  begin  with,  psychoanalytic  theory  assumed  that  all  psychologically 
relevant  structures  arise  in  ontogeny.  But  at  present  some  of  these 
structures  are  considered  to  be  congenitally  given.  This  shift  has  two 
implications :  first,  that  such  constitutionally  given  apparatuses  as  motility, 
perceptual  system,  memory  system,  thresholds50  are  psychologically 
relevant;  second,  that  the  ego  does  not  derive  from  the  id,  but  rather 
both  emerge  from  the  common  undifferentiated  matrix  of  the  first  ex- 
trauterine  phase  of  ontogenesis  [cf.  Hartmann,  Kris,  and  Loewenstein, 

While  originally  all  structures  were  considered  to  be  related  to  drive 
and  conflict,  it  is  now  assumed  that  the  inborn  ego  apparatuses  enter 
conflicts  as  independent  factors  and  that  their  function  is  not  primarily 
dependent  on  drives:  thus  they  are  termed  ego  apparatuses  of  primary 

48  Of  these  three  major  structural  concepts,  in  the  following  we  shall  discuss 
only  the  ego.  The  structural  treatment  of  the  id  and  superego  is  still  so  inadequate 
that  the  lengthy  discussion  it  would  require  is  beyond  the  scope  of  this  presentation. 

49  The  fact  that  certain  ego  structures   (e.g.,   defenses)   have  cognitive  repre- 
sentations does  not  contradict  the  distinction  made  here  between  defensive  struc- 
tures on  the  one  hand  and  cognitive  (means-)  structures  on  the  other. 

50  Cf .  note  47  above. 


autonomy.  [Cf.  Hartmann,  157;  see  also,  266,  280.]  This  does  not  imply 
that  they  have  no  relation  to  drives.  They  are  part  and  parcel  of  the  ap- 
paratus which  executes  drive  actions:  they  are  the  only  means  of  action 
the  organism  has.  Nor  does  it  imply  that  they  are  forever  free  of  conflict  : 
under  certain  conditions  they  can  and  do  become  involved  in  conflict, 
as  does  the  motor  apparatus  in  functional  paralyses  and  the  perceptual 
apparatus  in  the  tubular  vision  of  hysterics.  Thus  even  the  apparatuses  of 
primary  autonomy  are  only  relatively  autonomous  from  drive  and  con- 
flict. But  their  autonomy  does  imply:  first,  that  drives  only  trigger  their 
function  and  do  not  determine  their  course;  second,  that  they  can  and 
do  function  even  when  they  do  not  serve  the  gratification  of  a  specific 
drive.  Yet  reservations  must  be  made  on  both  of  these  points:  first, 
while  it  is  true  that  the  role  of  drives  in  relation  to  these  apparatuses  is 
primarily  that  of  triggering  their  function,  there  is  evidence  to  show  that 
the  drives  also  have  other  effects  on  the  apparatuses  (e.g.,  the  effects  of 
motivation  on  memory  [258] ) ;  second,  the  problem  of  the  energy  supply 
of  these  apparatuses  (when  they  are  not  triggered  by  drives)  has  so  far 
not  been  satisfactorily  solved  [161,  206,  and  268,  particularly  part  7]. 
Woodworth's  [329]  conception  of  "habit  as  drive,"  Buehler's  [42] 
"pleasure  in  functioning/'  Piaget's  [254]  "circular  reaction,"  Allport's 
[8]  "functional  autonomy"  imply  the  same  problem.  Attempted  solutions 
either  attribute  drives  (or  partial  drives)  to  apparatuses,  or  consider 
apparatuses  as  sources  of  (neutral)  ego  energy,  or  assume  that  the  energy 
they  use  is  neutralized  drive  energy  at  the  disposal  of  the  ego. 

Psychoanalytic  theory  at  first  considered  the  structures  which  arise 
in  the  course  of  ontogeny  as  conflict-born — i.e.,  defensive.  Since  defenses 
are  central  to  psychoanalytic  therapy,  they  are  the  most  extensively 
studied  structures,  and  this  gives  the  impression  that  all  structures  are 
conflict-born  and  all  controls  are  defenses.  Even  though  the  role  of 
identifications  in  building  ego  structures  was  recognized  early  [126],  a 
tendency  persists  to  consider  this  kind  of  structure-building,  too,  as 
conflict-born.  There  is  no  theoretical  clarity  even  now  on  this  point 
[see,  however,  Erikson,  66] :  certain  identifications  definitely  arise  from 
conflict  (e.g.,  identification  with  the  aggressor) ;  others  do  not  seem  to. 
But  it  is  clear  that  means-structures  born  in,  or  used  in,  the  course  of 
drive-gratification,  or  in  the  course  of  a  defensive  battle  against  drives, 
or  in  the  course  of  resolving  a  conflict  can  and  often  do  undergo  "change 
of  function"51  and  become  means  of  action  and  adaptation  in  the  service 
of  the  ego.  These  are  termed  structures  of  secondary  autonomy  [157, 
162].  They,  too,  are  only  relatively  autonomous  in  the  same  sense  as  are 
the  apparatuses  of  primary  autonomy.  They,  too,  are  assumed  to  have 

51  See  Hartmann  [157,  162];  for  instance,  rationalization  is  a  defense  mechanism 
which  tends  to  undergo  a  "change  of  function"  and  thus  to  become  an  important 
means  of  adaptation,  as  a  crucial  ingredient  of  logical  thought  and  rational  action. 

The  Structure  of  Psychoanalytic  Theory  97 

neutral  energies  at  their  disposal  or  to  use  neutralized  energies  at  the  dis- 
posal of  the  ego.  The  observations  to  which  the  concept  of  secondary 
autonomy  refers  are  akin  to  those  which  led  to  Allport's  concept  of 
"functional  autonomy" :  they  pertain  to  the  relation  of  means-structures 
to  basic  as  well  as  higher-level  motivations,  attitudes,  values,  etc. 

In  conclusion: 

1.  The  structural  determiners  of  behavior  were  introduced  as  inter- 
vening variables  to  account  for  the  observation  that  motivations  do  not 
determine  behavior  in  a  one-to-one  fashion. 

2.  Structural  determiners  differ  from  motivational  determiners  in 
that  they  are  relatively  permanent :  their  rate  of  change  is  relatively  slow. 

3.  There  are  inborn  structures  and  acquired  structures:  apparatuses 
of  primary  and  secondary  autonomy. 

4.  Structure-building  transforms  motivations  and  thus  gives  rise  to 
new  (more  neutralized)  motivations. 

5.  Structures  built,  and  the  motivations  arising  from  them,  may  be- 
come relatively  autonomous  determiners  of  behavior. 

I.  All  Behavior  Is  Determined  by  Reality  (the  Adaptive  Point  of 


Reality  in  psychoanalytic  theory  designates  the  external  source  of 
stimuli,  including  the  subject's  body,  but  excepting  the  somatic  sources 
of  drives  and  affects  [115,  pp.  60-64].  In  this  theory  reality  (i.e.,  external 
reality]  is  the  antithesis  of  psychological  reality™  [98,  pp.  548-549]. 

This  thesis  of  psychoanalytic  psychology  has  undergone  perhaps  more 
metamorphoses  than  any  other,  and  its  implications  are  far-reaching.  It 
implies,  from  the  point  of  view  of  psychology,  the  question  of  the  role 
stimuli  play  in  behavior;  from  that  of  biology,  the  question  of  the  rela- 
tionship between  the  organism  and  its  environment;  from  that  of  philos- 
ophy, the  epistemological  question  (i.e.,  how  man  can  know  of,  and  act 
in  accord  with,  his  environment  when  his  thoughts  and  actions  are  de- 
termined by  the  laws  of  his  own  nature) . 

In  psychoanalytic  theory's  first  conception,  reality  was  considered  the 
target  of  defense  [94,  ms.  H;  96,  97].  More  precisely,  the  defense  was 
directed  against  the  memory  of  a  real  event,  in  order  to  prevent  the  re- 
currence of  the  attendant  feelings  (affects)  which,  being  socially  pro- 
hibited, were  incompatible  with  self-respect™  Thus  symptoms  (i.e., 

52  The  corresponding  adjectives  are  real  and  psychological. 

53  Self-respect  and  social  prohibition  did  not  attain  conceptual  status  in  psycho- 
analysis proper.  They  cropped  up  in  the  early  superego  theory  [114],  and  returned 
in  the  Neo-Freudian  schools.  Only  recently  did  ego  psychology  begin  to  work  out 
their  place  in  the  theory  [Hartmann,  161;  Bibring,  27;  Erikson,  66;  Jacobson,  190, 


pathological  behavior)  resulting  from  defense  (repression,  etc.)  were 
considered  to  be  ultimately  determined  by  reality  events.  But  the  ques- 
tion of  reality's  role  in  determining  normal  behavior  was  not  yet  raised, 
although  it  was  assumed  that  the  affect  of  nontraumatic  experiences  is 
"dissipated33  by  being  distributed  over  a  wide  associative  network  [cf. 
35^  pp.  7_g;  'and  120]  while  the  affect  of  traumatic  experiences  is 

"dammed  up." 

The  second  conception  of  reality  [98],  which  dominated  psycho- 
analytic theory  from  1900  till  1923— with  the  exception  of  Freud's  "Two 
Principles"  [108],  which  prepared  the  next  conception— had  two  as- 
pects: these  were  the  drive  object  and  the  secondary  process. 

Drive  was  conceived  of  as  an  internal  stimulus  [115,  119]  which, 
unlike  external  stimuli,  is  continuous  and  inescapable  through  flight,  a 
stimulus  for  which  the  organism  has  no  intensity-reducing  barrier  of  the 
sort  which  operates  in  regard  to  external  stimuli.  In  turn  external  stimuli 
were  accorded  little  significance  and  psychological  effectiveness,  and  no 
behavior-determining  role.  Yet  at  the  same  time  certain  patterns  of  ex- 
ternal stimuli,  namely,  drive  objects,  were  conceived  of  as  the  precondi- 
tion for  drive  action  (drive  discharge).  Thus  the  effectiveness  of  drives, 
as  the  ultimate  determiners  of  behavior,  remained  in  part  dependent  on 
the  availability  of  the  drive  object.  Nor  is  this  the  only  role  reality  plays 
in  this  theory:  configurations  of  reality  which  prohibit  drive  action  were 
considered  to  be  represented  intrapsychically  by  the  censorship  [98]. 
This  is  a  drive-centered  conception  of  reality:  it  comprises  only  the 
conditions  which  make  drive  action  possible  or  impossible.  One  feature 
of  this  drive-object  conception  of  reality  has  a  broader  significance. 
While  the  instincts  of  animals  on  lower  evolutionary  levels  appear  to  be 
directly  and  more  or  less  rigidly  coordinated  to  specific  external  stim- 
uli, the  instincts  of  animals  on  higher  evolutionary  levels  appear  to  be 
less  rigidly  coordinated  to  such  specific  stimuli.  This  difference  may  be 
characterized  as  a  progressive  internalization  of  the  regulation  of  be- 
havior.54 The  psychoanalytic  theory  of  drives  assumes  that  the  relation 
of  human  drives  to  their  drive  objects  is  flexible,  and  that  the  regulation 
of  human  behavior  is  to  a  large  extent  internalized  [see  101].  Though 
early  psychoanalytic  theory  may  at  times  have  given  the  impression  that 
the  organism  is  totally  autonomous  from  its  environment,  it  was  never 
so  blind  as  to  take  this  extreme  stand.  But  it  certainly  does  raise  the 
question  of  the  organism's  relative  autonomy  from  its  environment  [cf .  pp. 
95-97,  above;  see  also  Gill  and  Brenman,  148;  and  280],  and  does  make 
it  clear  that  any  explanation  of  behavior  must  come  to  grips  with  the 
relative  autonomy  of  behavior  from  both  drives  and  external  reality. 

54  This  internalization  is  considered  coterminous  with  the  establishment  of  the 
ego;  cf.  Hartmann  [157,  160]. 

The  Structure  of  Psychoanalytic  Theory  99 

The  other  aspect  of  the  conception  of  reality  in  this  phase  of  psycho- 
analytic theory  was  the  secondary  process.  According  to  this  theory,  the 
secondary  thought  processes  tend  to  reflect  reality  "truly,"  i.e.',  not 
merely  in  terms  of  the  drive  whose  object  is  to  be  reached,  but  in  terms 
of  the  "actual"  relationships  between  objects  which  obtain  in  reality 
[98,  pp.  509,  533-536].  Likewise  the  secondary  processes  of  action  are 
"adaptive  to"  reality  [108,  pp.  15ff.]  and  do  not  strive  blindly  toward 
drive  discharge.  Delay  of  discharge,  detour  for  the  sake  of  a  safe  path, 
"full"  availability  of  memories  and  their  use  in  the  experimental  action 
of  thought  characterize  the  secondary  process,  which  is  thus  not  "selec- 
tive" in  the  limiting  sense  that  the  primary  process  is  but  has  a  broad 
access  to  reality  over  which  it  exercises  selective  judgments  and  choices. 
This  conception  implied  an  "objective"  reality,  and  secondary  processes 
which,  unlike  primary  processes,  do  not  "distort"  but  are  "veridical," 
even  though  it  was  clearly  recognized  that  the  secondary  process  cannot 
fully  reduce  these  "distortions"  because  to  do  so  it  would  have  to  elimi- 
nate the  affects  which  it  needs  as  its  orienting  signals  [98,  p.  536].  This 
conception  remained  incomplete,  since  it  left  the  origin,  nature,  and 
function  of  the  secondary  process  unexplained  [compare,  however,  Freud, 
129,  130;  also  Ferenczi,  78,  79,  80,  81]. 

The  third  conception  of  reality  appears  in  Freud's  ego  psychology,  of 
the  1923-1938  period,  and  was  forecast  in  the  "Two  Principles"  [108], 
particularly  by  the  concepts  of  reality  principle  and  reality  testing.  In  the 
first  conception,  the  defense  was  directed  against  reality  and  the  memory 
of  real  events.  In  the  second  conception,  it  was  directed  against  the  drive, 
and  reality  had  only  a  peripheral  role.  In  the  third  conception,  reality 
and  drive  appear  to  gain  a  more  or  less  equal  status  [131].  Now  the 
ultimate  motive  (determiner)  of  defense  is  real  danger,  and  the  drive 
is  defended  against  because  if  it  were  acted  upon  it  would  again  lead 
into  a  dangerous  real  situation.  Thus  defenses  against  drives  come  to 
represent  reality  and,  as  constituents  of  ego  and  superego  structure,  they 
become  internalized  regulations  of  behavior. 

In  the  period  we  are  now  considering,  the  ego  was  still  regarded  as 
a  mainly  defensive  organization;  nevertheless  its  origin  in  identifications55 
and  its  most  general  definition56  point  to  its  other  functions  and  to  its 
intimate  relation  to  reality.  The  identifications  with  the  objects  of  social 
reality  imply  that  reality  has  not  only  a  defensive-conflictful  role,  but 
also  an  ego-structure-forming  role.  Moreover  the  ego,  conceived  as  a 
cohesive  organization  with  a  synthetic  function  of  its  own  [131,  pp.  25- 
26;  245],  gains  a  degree  of  independence  from  drives  which  permits  a 

55  "The  ego  is  a  precipitate  of  abandoned  object  cathexes"  [126,  p.  36],  i.e.,  of 
identifications  with  abandoned  objects. 

sa  "^he  eg0]  is  a  coherent  organization  of  mental  processes"  [126,  p.  15]. 


relative  objectivity  in  regard  to  reality.  The  relationship  of  the  ego  to 
reality  brings  into  sharp  relief  the  central  role  of  reality  in  this  phase  of 
the  theory:  the  ego's  function  is  to  reconcile  the  demands  of  the  id, 
superego,  and  reality  [126];  the  relation  to  reality  is  crucial  to  the  ego 
[137];  and  the  ego  is  organized  around  the  System  Perception-Con- 
scious, i.e.,  around  the  means  of  contact  with  reality  [126], 

In  this  conception,  reality  shapes  not  only  the  ego,  but  even  the 
drives,57  which  were  previously  conceived  of  as  unchanging.  Moreover, 
in  Anna  Freud's  [93,  pp.  96,  109-110]  conception,  the  defense  against 
reality  itself  again  appears  as  a  concept,  much  as  it  appeared  in  Freud's 
first  conception  of  reality. 

The  fourth  conception  of  reality — Hartmann's — is  a  radical  develop- 
ment: the  organism,  as  a  product  of  evolution,  is  bom  already  adapted, 
or  potentially  adapted,  to  reality.  The  ego  apparatuses  of  primary 
autonomy  are  instruments  of  and  guarantees  of  man's  "preparedness  for 
an  average  expectable  environment."  In  animals  of  lower  evolutionary 
levels  the  instincts  are  the  guarantees  of  reality  adaptedness;  man's  drives 
have  lost  much  of  this  role,  and  thus  inborn  adaptedness  is  with  him 
more  a  potentiality  than  an  actuality;  processes  of  adaptation  outweigh 
inborn  adaptedness.  This  potentiality  for  internalized  regulation  of  be- 
havior actualizes  in  the  course  of  the  development  of  the  ego,  which  thus 
becomes  man's  organ  of  adaptation.58 

In  this  conception  reality  and  adaptedness  as  well  as  adaptation  to 
it  play  a  much  more  extensive  role  than  in  Freud's  third  theory  [cf .  Kris, 
205;  also  Hartmann,  165]:  here  they  are  the  matrix  of  all  behavior. 
Hartmann's  concepts  of  relative  autonomy.,  secondary  autonomy,  auto- 
matization, and  neutralization  for  the  first  time  provide  a  framework  for 
understanding  the  development  and  the  function  of  the  secondary  process 
as  one  of  man's  major  adaptative  means.  But  Hartmann  goes  even 
further  and  conceives  of  the  reality  to  which  man  adapts  as  one  created 
by  him  and  his  predecessors.  Yet  even  this  conception  seems  to  retain 
an  essential  duality  of  psychological  and  external  reality. 

The  fifth  conception  of  reality,  foreshadowed  by  both  Freud's  third 
conception  and  Hartmann's,  is  the  psychosocial  one  developed  by  Erikson 
[61,  66].  Man  is  potentially  preadapted,  not  only  to  one  average  expect- 
able environment,  but  to  a  whole  evolving  series  of  such  environments. 
These  environments  to  which  man  adapts  are  not  "objective,"  but  rather 
social  environments  which  meet  his  maturation  and  development  half- 
way: social  modalities  (e.g.,  the  socially  accepted  forms  of  "getting") 
foster,  select,  and  harness  his  developing  modes  (e.g.,  the  incorporative 

5J  See  [139],  but  note  this  point  already  in  [117]. 

58  Cf.  Hartmann:   Ego  Psychology  and  the  Problem  of  Adaptation  [157];  see 
also  [160]. 

The  Structure  of  Psychoanalytic  Theory  101 

oral  mode)  of  behavior  [62]. 59  This  is  the  genetic  counterpart  of  Hart- 
mann's  systematic  formulation;  it  is  thus  far  the  only  attempt  to  con- 
ceptualize the  phases  of  epigenesis  [58,  61]  through  which  preadapted- 
ness  becomes  effective,  and  in  which  processes  of  adaptation  inseparably 
unite  behavior  epigenesis  and  environmental  conditions  [61,  particularly 
chap.  7].  The  conceptions  of  an  "objective"  reality  and  of  an  unselective 
and  "veridical"  secondary  process  disappear  here  and  even  the  major 
(time  and  space)  coordinates  of  reality  become  "subjective"  [63],  in  the 
sense  that  they  are  shown  to  be  relative  both  to  organismic  space  and 
time  and  to  the  particular  society  into  whose  reality  man  grows.60 

J.  All  Behavior  Is  Socially  Determined  (the  Psychosocial  Point  of 


To  demonstrate  that  psychoanalysis  considers  real  experience  in 
general  and  social  experience  in  particular  to  be  determiners  of  behavior 
is  to  bang  on  an  open  door.  For  instance,  organic  psychiatry,  which 
centered  on  constitutional  and  hereditary  factors,  has  always  regarded 
psychoanalysis  as  a  pure  "nurture  psychology,"  but  to  do  so,  it  had  to 
disregard  the  "nature"  conception  of  drives  in  psychoanalytic  theory.  The 
root  of  this  and  kindred  misunderstandings  seems  to  be  that  the  theory 
did  not  systematically  clarify  its  stand  on  the  dual  relationship  between  the 
organism  and  its  environment.  It  is  characteristic  of  organisms  that  they 
are  dependent  on  their  environment  but  also  relatively  independent  from 
it.  This  balance  between  dependence  and  independence  might  be 
designated  as  a  relative  autonomy  (of  the  organism  from  its  environ- 
ment) in  the  same  sense  in  which  the  analogous  relationship  of  the  ego 
to  the  id  is  so  designated  [280]. 

The  organic  psychiatrists'  view  implies  an  absolute  autonomy  from 
environmental  influences.  From  their  point  of  view,  Freud  completely 
disregarded  this  autonomy,  since  he  dealt  mainly  with  the  dependence 

69  In  this  conception  modes  develop  according  to  genetic,  inborn  laws,  but  the 
social  organization  of  the  environment  defines  their  place  and  form  in  the  be- 
havior repertory  and  their  use  in  reality  mastery  and  adaptation. 

60  This  conception  does  not  deny  the  "objectivity"  of  the  common,  consensually 
validated  aspects  of  space  and  time,  or  the  intellectual  possibility  of  transcending 
the  subjective  coordinates  of  reality  in  order  to  build  universally  valid  sciences  of 
space,  matter,  etc.  It  does  not  invalidate  the  coordination  of  the  organism  with 
the  "objective"  environment,  which  is  guaranteed  by  the  apparatuses  of  primary 
and  secondary  autonomy  (e.g.,  the  perceptual  and  motor  apparatuses),  nor  the 
effectiveness  of  the  "causal  texture"  of  the  environment  (Brunswik,  Heider)  which 
sets  limits  to  all  individual  and  social  "subjectivity."  For  a  detailed  discussion  of 
these  issues,  which  lead  far  into  perception  theory,  see  G.  Klein's  forthcoming 
volume  [198]. 


of  behavior  on  experience,  societal  norms,  etc.  On  the  other  hand,  Freud's 
stress  on  drives,  as  well  as  Ms  regard  for  constitutional  factors,  made 
his  theory  appear — even  in  the  eyes  of  many  of  his  followers — to  imply 
absolute  autonomy  from  the  environment.  From  this  vantage  point 
society  appeared  as  a  factor  interfering  with  nature,  man  was  looked 
upon  as  a  born  individualist,  and  therefore  the  therapeutic  aim  was 
often  regarded  as  the  liberation  of  human  nature  from  social  bondage. 

Ho\vever,  Freud  considered  the  sexual  drives  and  their  object  choices 
as  anaclitic61  upon  the  drives  of  self-preservation  and  their  object  choices. 
This  is  a  statement  of  the  growing  organism's  first  social  relationships 
and  implies  the  social  determination  of  behavior.  So  does  the  conception 
of  the  Oedipus  complex:  the  budding  individual's  social  environment 
provides  the  objects  of  his  libidinal  and  aggressive  drives,  and  the 
structures  (identifications  in  ego  and  superego  development)  which 
the  relationships  between  the  subject  and  these  objects  give  rise  to, 
codetermine  his  behavior  in  general  and  not  merely  his  pathology  [126]. 
Though  these  social  conceptions  were  not  generalized  into  an  explicit 
psychoanalytic  social  psychology,  the  social  determination  of  behavior 
is  clearly  not  alien  to  classic  psychoanalytic  theory. 

Why  then  the  reluctance  of  classic  psychoanalysis  to  accept  the 
emphasis  placed  on  the  social  determination  of  behavior  by  Adler, 
Sullivan,  Homey,  and  Kardiner?  It  seems  that  a  struggle  between  dif- 
ferent conceptions  of  the  relative  autonomy  of  behavior  from  environ- 
mental reality62  lay  behind  this  reluctance.  To  the  classic  analyst's 
mind,  the  "dissident"  schools,  upon  discovering  the  dependence  of  be- 
havior on  social  reality,  abandoned  those  concepts  of  the  psychoanalytic 
theory  which  encompassed  the  observations  concerning  the  autonomy 
of  behavior  from  the  environment:  these  were  the  drives  and  the  other 
constitutional  (e.g.,  structural)  givens.  The  net  result  seemed  to  be  that 
some  dissidents  came  to  regard  adaptation  as  "adjustment"  (particularly 
as  a  therapeutic  goal),  to  disregard  the  existence  and  nature  of  drives, 
to  stress  the  environmental  demand,  and  thus  deliberately  or  unwittingly 
to  reinforce  censorship  and  superego.  These  therapists  were  said  to  have 
come  to  take  "society's  side"  against  the  patient — although  their  task, 
as  originally  conceived,  was  to  take  neither  or  both.  In  other  dissidents 
the  result  seemed  to  be  very  different:  society  was  blamed  for  man's 
troubles  and  was  to  be  so  changed  as  to  cause  no  more  trouble  to  man. 
This  was  said  to  be  a  stand  on  the  side  of  the  individual  against  society. 
Thus,  do-goodism,  social  rebelliousness,  Philistine  demands  for  con- 

61  Anaclitic:    leaning  upon.  The  implication  is  that   the  first   objects   of  the 
sexual  drive  are  the  people  who  take  care  of  the  infant  and  guarantee  his  survival, 
i.e.,  who  are  the  objects  of  his  self-preservative  drives  [101], 

62  Gf.5  for  this  section,  Fenichel's  discussion  of  Fromm  [76]. 

The  Structure  of  Psychoanalytic  Theory  103 

formity,  and  attitudes  so  approving  of  individual  liberty  as  to  shade 
into  license  were  considered  to  be  characteristic  of  the  dissidents.  Any 
of  these  attitudes  may  well  be  a  therapist's  private  convictions,  and 
as  such  cannot  but  enter  therapy  in  some  form,  but  none  of  them  has  a 
place  in  psychological  theory.  It  is  not  our  task  to  establish  whether 
these  are  fair  assessments  of  the  dissident  schools.63  It  is  sufficient  to 
point  out  that  these  imputed  or  actual  attitudes  are  indices  of  a  struggle 
centering  around  the  ego's  relative  autonomy  from  social  reality.  The 
dissidents'  contribution  toward  the  formulation  of  the  psychosocial  point 
of  view  must  not  be  underestimated. 

Anna  Freud's  stress  on  work  with  the  parents  of  her  child  patients, 
and  her  work  with  groups  of  children  in  the  course  of  the  war,  is  a 
recognition  of  the  social  determination  of  behavior. 

Hartmann  gave  the  first  theoretical  formulation  of  the  role  of  social 
reality.  His  point  of  departure  was  Freud's  conception  of  the  central 
role  in  human  development  of  the  infant's  prolonged  helplessness  and 
dependence  on  caretaking  adults.  Hartmann's  major  theoretical  advance 
is  embodied  in  his  concept  of  "social  compliance,"  which  is  coined  on 
Freud's  concept  of  "somatic  compliance."  The  referents  of  this  concept 
are  the  observations  concerning  those  institutions  of  society  which  meet, 
foster,  and  mold  the  developing  individual's  inborn  and  acquired 
adaptive  means.  Education  as  a  social  institution  is  an  instance  of  "social 
compliance"  [157], 

Erikson's  epigenetic  psychosocial  conception  parallels  and  fills  in 
Hartmann's  systematic  and  programmatic  formulations.  Erikson's  con- 
ception of  society  is  detailed :  it  is  the  geography  and  the  means  of  sur- 
vival it  provides;  it  is  the  economy  and  other  social  institutions;  it  is  the 
ideology,  including  tradition  [61,  chaps.  3,  4;  also  60].  It  involves:  (1) 
the  epigenetic  conception  of  ego  development  consisting  of  a  sequence  of 
developmental  phases,  each  characterized  by  a  phase-specific  crisis  which 
is  universal,  while  its  solution  varies  from  society  to  society  and  is  in- 
dividually unique  [61,  chap.  2;  62,  66];  (2)  the  caretaking  people 
(parents,  teachers,  etc.)  and  their  practices  representing  the  society's  in- 
stitutions and  traditions  which  were  developed  to  meet  each  phase-specific 
crisis  of  the  developing  individual's  life  cycle;  (3 )  the  phase-specific  needs 
of  the  growing  individual  eliciting  dovetailing  needs  in  the  caretaking 
people,  which  correspond  to  the  respective  phases  of  their  life  cycles ;  the 
society's  institutions  and  traditions  making  their  contribution  to  the 
solution  of  the  growing  individual's  phase-specific  crisis  by  means  of  the 
so-elicited  needs  of  the  caretaking  people;  (4)  the  resulting  behavior 

63  FenicliePs  assessment  [76]  of  Kardiner  and  Fromm  seems  as  adequate  as  the 
conceptual  equipment  of  the  time  permitted:  Hartmann's  and  Erikson's  concepts 
had  not  yet  entered  the  argument. 


forms  having,  by  and  large,  an  accepted  place  in  the  society  and  guaran- 
teeing the  individual's  viability  in  it. 

In  Erikson's  conception  neither  does  the  individual  adapt  to  society 
nor  does  society  mold  him  into  its  pattern;  rather,  society  and  individual 
form,  a  unity  within  which  a  mutual  regulation  takes  place.  The  social 
institutions  are  preconditions  of  individual  development,  and  the  develop- 
ing individual's  behavior  in  turn  elicits  that  help  which  society  gives 
through  its  adult  members  directed  by  its  institutions  and  traditions. 
Society  is  not  merely  a  prohibitor  or  provider;  it  is  the  necessary  matrix 
of  the  development  of  all  behavior.  Indeed,  the  development  and 
maintenance  of  the  ego,  of  the  superego,  and  perhaps  of  all  structures 
are  dependent  on  the  social  matrix :  behavior  is  determined  by  it  and  is 
possible  only  within  it. 

BetteUieim's  [21,  22,  23]  and  Redl's  [282,  283]  studies  confirmed 
this  conception  and  extended  it.  Gill  and  Rapaport  [149]  concluded, 
from  the  observations  and  theories  here  discussed,  that  the  meta- 
psychological  triad  of  the  dynamic,  structural,  and  economic  points  of 
view  must  be  extended  by  the  addition  of  an  adaptive  point  of  view. 

K.  Discussion 

This  sketch  of  the  basic  propositions  of  the  general  psychoanalytic 
theory  was  presented  to  make  a  discussion  of  its  variables  possible.  It 
centered  around  the  three  classic  metapsychological  points  of  view 
(dynamic,  topographic,  and  economic)  [cf.  117,  p.  114],  but  it  also  in- 
cluded the  structural  point  of  view  (which  elaborated  and  replaced  the 
topographic  one)  as  well  as  the  genetic  and  the  adaptive  points  of  view, 
which  (being  of  the  same  order  of  significance  in  the  theory  as  the  classic 
triad)  seem  necessary  to  complete  the  system  of  psychoanalytic  metapsy- 
chology  [see  Gill  and  Rapaport,  149].  The  inclusion  of  the  psycho- 
social  point  of  view  (like  that  of  the  topographic  one)  is  a  mark  of 
systematic  weakness,  since  it  is  merely  a  specific  aspect  of  the  adaptive 
point  of  view.  It  is  as  yet  difficult,  if  not  impossible,  to  present  the 
theory  divorced  from  its  history.  The  topographic  point  of  view,  though 
it  is  satisfactorily  replaced  by  the  structural  one,  appears  here  because  it 
is  difficult  to  present  the  latter  so  that  the  role  of  unconscious  deter- 
mination will  emerge  as  clearly  as  it  does  from  the  by  now  historical 
topographic  point  of  view.  Likewise,  the  psychosocial  point  of  view  is 
discussed  separately  because  it  is  as  yet  difficult  to  present  the  adaptive 
point  of  view  so  that  its  psychosocial  implications  emerge  clearly.  Both 
Hartmann's  and  Erikson's  theories  are  too  new,  their  implications  too 
little  understood,  and  their  relationship  to  each  other  too  little  ex- 
plored [see  Rapaport,  277,  278]  to  permit  a  statement  disregarding  all 
but  systematic  considerations. 

The  Structure  of  Psychoanalytic  Theory  105 

Thus  it  is  likely  that  seven  of  these  ten  points  of  view  which  we 
have  discussed  here  will,  in  future  systematic  treatments,  be  condensed 
into  the  five  (dynamic,  economic,  structural,  genetic,  and  adaptive) 
metapsychological  points  of  view.  Indeed,  they  may  eventually  be 
formulated  as  the  axioms  of  the  system.  The  remaining  three  points  of 
view  (empirical,  Gestalt,  and  organismic)  seem  to  be  of  a  different 
character,  and  lumping  them  together  with  the  metapsychological  points 
of  view  is  another  indication  that  the  systematization  here  attempted  is 
premature.  By  and  large,  they  deal  with  the  theory's  observables  and 
with  their  organization  as  units.  It  is  possible  that  these  three  points  of 
view  will  appear,  in  future  systematic  statements,  as  definitions  of  observ- 
ables. If  so,  then  why  were  they  not  segregated  here  from  the  others? 
One  reason  is  that  the  others  are  not  yet  formulated  as  axioms  and  they, 
too,  imply  definitions.  Then  why  not  explicate  all  the  definitions  and 
segregate  them  from  the  points  of  view?  At  this  stage  of  our  knowledge, 
even  if  such  explication  and  segregation  were  possible,  the  present  analysis 
would  be  an  indispensable  preparatory  step. 

Tolman's  [316],  MacGorquodale  and  Meehl's  [227],  and  others' 
conceptions  of  independent,  intervening,  and  dependent  variables  can- 
not be  discussed  here  in  detail.  It  should  suffice  to  say  that  their  views 
of  these  variables  do  not  seem  to  be  "methodologically  pure,"  but 
rather  loaded  with  their  respective  systematic  biases.63a  Only  that  con- 
ception of  variables  which  Koch  calls  the  "mathematical"  seems  relevant 
to  psychoanalytic  theory.  It  is  in  the  sense  of  such  a  mathematical  con- 
ception that  we  will  speak  here  of  variables. 

I  believe  that  the  following  conclusions  may  be  derived  from  the 
sketch  of  the  theory's  "points  of  view" : 

1.  The  psychoanalytic  concept  of  over  determination  implies  that  one 
or  several  determiners  of  a  given  behavior,  which  appear  to  explain  it, 
do  not  necessarily  give  its  full  causal  explanation.  This  is  not  per  se  alien 
to  other  sciences,  though  a  principle  of  over  determination  did  not  be- 
come necessary  in  any  of  them.  Psychoanalysis'  need  for  this  principle 
seems  to  be  due  partly  to  the  multiplicity  of  the  determiners  of  human 
behavior,  and  partly  to  the  theory's  characteristic  lack  of  criteria  for  the 
independence  and  sufficiency  of  causes.  The  determiners  of  behavior  in 
this  theory  are  so  defined  that  they  apply  to  all  behavior  and  thus  their 
empirical  referents  must  be  present  in  any  and  all  behavior.  Since  there 
is  usually  no  single  determiner  which  constantly  assumes  the  dominant 
role  in  a  given  behavior,  other  determiners  can  hardly  be  neglected 
while  a  dominant  determiner  is  explored.  When  favorable  conditions 
make  one  determiner  dominant,  the  investigator  is  tempted  to  conclude 

63a  Frenkel-Brunswik  [92,  pp.  307ff.]  gives  a  cogent  discussion  of  some  of  these 
biases.  ' 


that  he  has  confirmed  a  predicted  functional  relationship — as  he  indeed 
has.  Regrettably,  the  attempt  to  repeat  the  observation  or  experiment 
in  question  often  fails,  because  in  the  replication  either  the  same  be- 
havior appears  even  though  a  different  determiner  has  become  dominant, 
or  a  different  behavior  appears  even  though  the  same  determiner  has 
remained  dominant.64  Lewin's  Gesetz  und  Experiment  in  der  Psychologic 
[215]  is  relevant  here:  it  argues  that  the  criterion  of  validity  for  psy- 
chological experiments  [cf.  Gill  and  Brenman,  34,  147;  and  Benjamin, 
1 1]  is  not  repeatability,  but  predictable  systematic  variation. 

The  implications  of  the  concept  of  overdetermination  for  the  choice 
of  independent  variables  are:  (a]  Any  motivation  high  in  the  hierarchy 
of  psychological  organization,  if  chosen  as  the  independent  variable  of 
an  experiment  or  observation,  may  prove  to  be  dependent  on  variables 
closer  to  the  base  of  the  hierarchy.  In  this  case,  either  the  dependent 
variable  will  be  treated  as  an  implicit  function  of  these  more  basic  varia- 
bles (motivations,  structures,  etc.),  or  the  latter  will  be  considered  as  inter- 
vening variables  interposed  between  the  independent  and  the  dependent 
variable,  (b)  If  a  basic  motivation  is  chosen  as  the  independent  variable 
of  an  experiment,  then  variables  higher  in  the  hierarchy  will  be 
interposed  as  intervening  variables  between  the  experiment's  independent 
and  dependent  variable.  For  instance,  in  Klein's  [197]  experiment, 
cognitive  attitudes  are  the  intervening,  thirst  the  independent,  and 
cognitive  behaviors  the  dependent  variables.  Though  in  this  theory  basic 
drives  are  systematically  distinguished65  independent  variables,  as  em- 
pirical independent  variables  they  do  not  seem  to  differ  significantly  from 
other  motivational  variables. 

2.  The  psychoanalytic  conception  of  autonomy  puts  a  further  limita- 
tion on  the  distinguished  independent  variable  character  of  basic  drives 
by  pointing  to  other  equally  distinguished  ones.  The  concept  of  autonomy 
implies  that  structures  of  primary  (and  secondary)  autonomy  may  re- 
tain (or  attain)  a  relative  independence  from  drives.  The  function  of 
autonomous  structures,  even  when  triggered  by  drives,  may  remain  in- 
dependent from  them.  Derivative  motivations  (as  a  rule,  related  to 

64  For  instance,  in  a  Zeigarnik-like  experiment,  on  the  one  hand  an  interrupted 
task   may   be   remembered    either   because    of   the   undischarged    tension   system 
(Lewin's  explanation)   or  because  the  task  had  a  specific  "historical"  or  "motiva- 
tional" significance  for  the  subject  to  begin  with.  In  this  instance  different  domi- 
nant determiners  have  identical  effects.  On  the  other  hand,  an  interrupted  task 
may  be  forgotten  (in  spite  of  the  undischarged  tension  system)   when  the  inter- 
ruption is  experienced  as  a  failure.  In  this  instance  the  undischarged  tension  re- 
mains the  dominant  determiner,  but  its  behavioral  effect  is  different. 

65  The  term   "distinguished"  is  used  here  to   convey  that  though  the  theory's 
development  placed  restrictions  on  the  initial  conception  of  drives  as  "ultimate 
causal   determiners,"   drives   still  retain  a  special  position  in  the  system  of  the 
theory  (see  pp.  89-91,  above). 

The  Structure  of  Psychoanalytic  Theory  107 

structures)  may  also  attain  such  independence.  The  implications  of 
this  concept  of  autonomy,  for  the  choice  of  independent  variables,  are 
contrary  to  those  of  overdetermination :  since  autonomy  increases  with 
the  distance  from  the  basic  drives,  variables  (structures  or  motivations) 
high  in  the  hierarchy  also  appear  as  systematically  distinguished  in- 
dependent variables.  As  independent  variables  of  an  experiment,  they 
have  the  advantage  that  the  dependent  variable  need  not  be  an  implicit 
function  of  more  basic  ones,  and  that  (in  the  ideal  case)  no  intervening 
variables  are  interposed  between  them  and  the  dependent  variable. 

The  conflicting  implications  of  the  "overdetermination"  and  the 
"autonomy"  concepts  limit  the  advantages  of  the  latter  also,  since  not 
all  structures  and  derivative  motivations  retain  or  attain  autonomy;  and 
moreover,  autonomy  is  not  an  all-or-nothing  affair,  but  rather  a  matter 
of  degree,  and  thus  we  always  deal  with  relative  autonomy,  the  degree 
of  which  must  be  empirically  determined. 

3.  Basic  motivations  and  structures,  as  well  as  motivations  of  a  high 
degree  of  autonomy,  are  systematically  distinguished  variables.  Whether 
they  should  be  considered  to  be  systematic  independent  variables  (in 
Koch's  sense)  is  not  clear,  since  they  may  also  appear  in  the  role  of  inter- 
vening and  dependent  variables.  To  illustrate  this,  let  us  survey  the  main 
classes  of  the  theory's  variables:  motivations  and  structures  (of  any 
hierarchic  level  and  degree  of  autonomy),  behaviors  (including  thought 
and  affect  as  well  as  observable  action),  and  external  reality. 

External  reality.  In  the  reflex-arc  model  external  reality  (stimulus) 
appears  as  the  independent  variable.  The  model  assumes  that  in  this 
case  unconscious  impulses  and  ideas  always  enter  as  intervening  variables, 
and  that  the  dependent  variable  is  motor  action  and/or  conscious 
thought  and/or  affect.  However,  the  autonomy  concept  implies  that 
the  functional  relationship  between  stimulus  and  behavior  may  be  of  any 
degree  of  relative  autonomy;  i.e.,  the  extent  to  which  unconscious  im- 
pulses and  ideas  intervene  may  vary.  Thus  S-R  psychology  appears 
here  as  a  limiting  case  of  a  high  degree  of  autonomy  (automatization). 
[See  Hartmann,  157,  pp.  26,  86ff.] 

External  reality  as  an  intervening  variable  is  one  of  the  implications 
of  the  adaptive  point  of  view.  When  either  structure  or  motivation  is 
chosen  as  the  independent,  and  behavior  as  the  dependent  variable, 
external  reality  appears  as  the  intervening  variable,  and  corresponds  to 
the  adaptive  aspect  of  the  behavior  in  question.66  The  concept  of  relative 
autonomy  from  the  environment,  however,  implies  that  some  of  the 
motivation  vs.  behavior  and  structure  vs.  behavior  relationships  (like 
those  in  impulsive  actions,  fugues,  and  characterologically  typical  be- 

66  For  instance,  in  studying  the  effect  of  hunger  on  feeding  behavior,  the  absence 
of,  or  presence  and  demeanor  of,  an  observer  will  enter  as  an  intervening  variable. 


haviors   respectively)    will   be,   within   limits,   invariant   in   regard   to 
changes  in  external  reality. 

How  external  reality  can  be  a  dependent  variable  is  less  obvious.  If 
external  reality  were  conceived  "geographically"  [see  Koffka,  204,  pp. 
27ff.]5  then  it  would  be  meaningless  (or  a  subject  matter  for  physics, 
chemistry,  etc. )  to  treat  it  as  a  dependent  variable.  If,  however,  we  go  be- 
yond its  inherent  "causal  texture"  (Brunswik,  Heider)  and  conceive  of  it 
psychologically,  then  it  can  become  a  dependent  variable.  A  person  in 
my  life  space  is  an  external  reality,  yet  this  external  reality  is  a  variable 
dependent  on  my  "feelings"  toward  this  person.  In  Bruner  and  Good- 
man's experiment,  the  sizes  of  coins  appear  as  dependent  variables, 
though  in  that  instance  it  is  difficult  to  separate  perceptual  behavior  and 
external  reality  as  dependent  variables. 

Motivations.  In  deprivation  experiments,  as  well  as  in  observations 
made  in  therapy  (e.g.,  on  transference  phenomena),  motivations  appear 
as  independent  variables  ;&7  and  their  hierarchic  position  (implying  con- 
siderations of  overdetermination  and  autonomy)  defines  the  degree  of 
their  actual  independence.  In  these  instances  defenses  and  other  struc- 
tures usually  enter  as  intervening  variables,  and  behavior  is  the  most 
common  dependent  variable,  though  in  investigations  concerning  proc- 
esses of  structure  building  and  structural  change,  structure  will  be  the 
dependent  variable. 

In  the  reflex-arc  model,  unconscious  motivations  appear  as  inter- 
vening variables,  and  external  reality  plays  the  role  of  the  independent, 
and  behavior  that  of  the  dependent  variable.  Moreover,  wherever  a  con- 
trolled (not  impulsive)  motivation,  or  a  structure  close  to  the  base  of  the 
hierarchy  of  the  mental  organization,  is  taken  as  the  independent 
variable,  higher-level  motivations  appear  as  intervening  variables,  pro- 
vided that  no  automatized  relationships  obtain  between  the  independent 
and  the  dependent  variables. 

Motivations  as  dependent  variables  are  encountered  when  motiva- 
tions close  to  the  base  of  the  hierarchy  are  chosen  as  independent  vari- 
ables and  defensive  structures  appear  as  intervening  variables;  or  when 
external  reality,  in  the  form  of  deprivation,  is  the  independent  variable. 
Clinically  the  presence  of  defensive  structures  is — as  a  rule — inferred 
from  the  appearance  of  derivative  motivations,68  which  are  in  this  case 
dependent  variables.  But  certain  motivations  may  also  appear  as  de- 
pendent variables  where  external  realities  (other  than  deprivation) 
or  psychological  structures  are  the  independent  variables. 

6T  Except  where  the  degree  of  deprivation  is  the  independent,  and  the  drive  the 
dependent  variable. 

68  For  instance,  in  studying  a  coprophilic  impulse,  the  presence  of  reaction 
formation  may  be  inferred  from  the  appearance  of  a  motivation  for  excessive 

The  Structure  of  Psychoanalytic  Theory  109 

Structures.  Structures  appear  as  independent  variables  wherever  in- 
dividual differences  in  behavior,  under  (relatively)  constant  motivation 
and  stimulation,  are  studied:  for  instance,  in  the  comparative  study  of 
symptoms  in  various  neuroses,  and  in  the  studies  of  individual  differences 
in  perception  [Klein,  Holzman,  Gardner,  Schlesinger,  144,  179,  180, 
198,  304]. 

Structures  as  intervening  variables  are  commonplace  in  clinical  ob- 
servation. They  account  for  the  lack  of  a  one-to-one  relationship  between 
motivations  and  behavior.  Defensive  structures  countermand  motivations 
and  replace  them  by  derivative  motivations  (as,  for  instance,  in  reaction- 
formation).  Controlling  structures  direct  and  channel  motivations,  as 
in  delay  and  detour-behavior  and  in  the  choice  of  substitute  goals.  In 
Klein's  [197]  thirst  experiment,  the  thirst  motivation  was  the  inde- 
pendent, perceptual  behavior  the  dependent  variable,  and  structures 
(the  subjects5  "cognitive  attitudes")  appeared  as  intervening  variables. 

It  is  less  easy  to  conceive  of  structures  as  dependent  variables, 
though  they  do  appear  as  such  in  processes  of  structural  change,  includ- 
ing those  of  learning.  In  so  far  as  psychoanalysis  as  therapy  achieves  its 
goal  of  changing  existing  structures,  in  at  least  some  of  the  observations 
made  in  therapy,  structures  appear  as  dependent  variables.  Piaget's 
[254]  schemata  of  the  primary,  secondary,  and  tertiary  circular  reactions 
are  structures,  and  in  his  developmental  studies  which  trace  their  growth 
and  fate,  structures  are  dependent  variables. 

Behavior.  The  role  of  behavior  as  a  dependent  variable  needs  no 
discussion.  But  it  might  be  worth  noting  again  that  here  behavior  is 
broadly  defined  to  include  -conscious  and  unconscious  thought,  affect, 
and  action,  which  can  and  do  substitute  for  one  another,  so  that  behavior 
is  a  complex  dependent  variable. 

The  role  of  behavior  as  an  intervening  variable  is  more  difficult  to 
conceive  of,  though  it  is  commonly  enough  encountered.  When,  for  ex- 
ample, a  motivation  is  taken  as  the  independent  variable  and  the  ob- 
servable action  facet  of  behavior  is  taken  as  the  dependent  variable, 
the  thought  and  affect  facets  of  behavior,  as  a  rule,  interpose  them- 
selves as  intervening  variables.  This  seems  to  be  one  of  Hebb's  [169] 
points  in  his  criticism  of  S-R  theories.  Naturally,  in  impulsive  actions 
and  where  the  relation  between  motivation  and  overt  action  is  auto- 
matized, such  intervening  variables  are  likely  to  be  absent. 

The  conception  of  behavior  as  an  independent  variable  is  perhaps 
the  least  obvious  of  all.  Yet,  for  example,  under  conditions  of  a  high 
degree  of  autonomy,  one  facet  of  behavior  may  be  taken  as  an  in- 
dependent and  any  other  facet  of  it  as  the  dependent  variable.  For  in- 
stance, in  Werner's  experiments  [152,  208] — in  which  the  subject 
presses  against  motor  restraint,  with  the  consequence  that  the  number 
of  his  movement  responses  in  the  Rorschach  test  increases — the  motor 


facet  of  behavior  is  the  independent,  while  its  cognitive  facet  is  the 
dependent,  variable. 

4.  It  seems  that  the  variables  of  all  these  major  classes  can  be  treated 
as  empirical  independent,  intervening,   and  dependent  variables;  but 
they  differ  greatly  in  regard  to  manipulability,  which  is  considered  by 
some  to  be  the  criterion  for  the  selection  of  independent  variables. 
Genetic  and  structural  variables,  for  instance,  are  not  amenable  to  direct 
manipulation.  Besides  such  intrinsic   difficulties,   manipulation  of  the 
internal  and  external  environments  as  well  as  of  action  is  also  limited 
by  due  regard  for  the  subjects'  privacy  and  by  the  fact  that  social 
manipulation  beyond  a  narrow  range  is  likely  to  endanger  the  in- 
dividual's rights.  But  manipulability  is  not  an  indispensable  criterion; 
it  may  be  replaced  by  observation  (as  in  astronomy),  or  by  seeking  out 
"nature's  experiments"  (as  in  evolution  theory) . 

5.  We  may  conclude  that  psychoanalytic  theory  requires  the  ex- 
ploration of  all  the  possible  functional  relationships  among  its  variables. 
One  wonders  whether  or  not  there  is  an  intimate  connection  between  the 
rigid  decisions  of  various  schools  of  psychology  on  systematic  variables 
(e.g.,  those  of  S-R  and  Gestalt  psychology)   and  the  limited  range  of 
observables  acceptable  to  each  of  them.  Any  limitation  on  the  choice 
of  variables  seems  to  result  in  a  limited  range  of  observables  and  ob- 
servational methods,  and  it  is  the  dearth  of  methods  which  is  probably 
the  major  obstacle  to  bridging  the  gap  between  psychoanalysis   and 
academic  psychology  [cf.  Shakow  and  Rapaport,  309],  and  between  the 
various  schools  of  psychology. 


A.  Initial  Evidential  Grounds 

We  will  discuss  here  only  the  evidential  grounds  for  the  early  as- 
sumptions of  the  system;69  to  trace  those  of  its  present  assumptions  would 
be  a  historical  job  far  exceeding  the  scope  of  this  essay.  Thus,  the 
propositions  to  be  discussed  in  this  section  are  not  always  identical  with 
those  of  the  present  theory. 

The  basic  assumption  of  psychoanalytic  theory  was  and  is  thorough- 
going psychological  determinism.  Its  other  initial  assumptions  are  im- 
plicit in  the  thesis  of  psychoanalytic  metapsychology:  a  full  description 
of  any  psychological  phenomenon  must  include  its  dynamic,  topographic, 

69  The  reference  here,  if  not  otherwise  indicated,  is  to  Breuer  and  Freud  [35]. 

The  Structure  of  Psychoanalytic  Theory  111 

and  economic  descriptions.70  What  are  the  assumptions  implied  in  these 
three  points  of  view? 

The  topographic  point  of  view  distinguishes  between  the  Systems 
Unconscious,  Preconscious,  and  Conscious,  and  thus  implies  the  as- 
sumption of  unconscious  psychological  processes,,  which  is,  except  for 
psychological  determinism,  the  earliest  and  most  general  assumption  of 
psychoanalysis.  On  this  assumption  are  built  the  concepts  of  unconscious 
motivation  (invoking  dynamic  assumptions)  and  primary  process  (in- 
volving economic  assumptions). 

The  dynamic  point  of  view  implies  the  assumption  of  psychological 
forces  and  their  conflicts  in  general  and  of  the  drive  nature  of  these 
forces  in  particular  [98].  On  these  assumptions  are  built  the  concepts  of 
libidinal  drives  and  censorship  (ego  drives  or  self-preservative  drives — 
which  are  now  dated),  as  well  as  the  conception  of  the  central  role  of 
libidinal  drives  [101]. 

The  economic  point  of  view  implies  the  existence  of  psychological 
energies  in  general  and  their  drive  origin  in  particular.  These  assump- 
tions underlie  the  concept  of  cathexis  (quantity  of  energy).  Corollaries 
of  these  assumptions  are  principles  analogous  to  the  physical  principles 
of  conservation  of  energy,  entropy,  and  least  action.  The  conservation 
principle:  cathexis  is  never  lost  and  thus  is  traceable  in  the  expenditures 
and  transformations  of  cathexes  involved  in  psychological  forces  [117, 
p.  114].  The  entropy  principle  (the  much  misconstrued  pleasure 
principle) :  drive  energy  tends  toward  discharge  (i.e.,  diminution  of 
tension)  [35,  p.  143;  and  98,  pp.  508-509,  533-535].  The  principle  of 
least  action:  processes  involving  cathexes  other  than  those  of  basic 
drives  operate  so  as  to  expend  the  least  amount  of  cathexis  [98,  pp.  533- 
534].  The  main  concepts  built  on  these  principles  are  wish-fulfillment 
vs.  reality-testing  which  direct  and  the  primary  vs.  secondary  process 
mechanisms  which  subserve  the  transfer  and  transformation  of  cathexes 
[98,  pp.  530-531,  535-536]. 

The  initial  evidence  for  these  three  sets  of  assumptions  and  their 
corollaries  cannot  at  this  time  be  sharply  separated  from  the  evidence 
for  the  validity  of  the  theories  built  upon  them.  A  sharp  separation  would 
require  prior  decisions  as  to  which  assumptions  are  to  be  treated  as 
axioms  and  which  are  to  be  empirically  derived  from  a  combination 
of  axioms,  definitions,  and  observations. 

1.  The  assumption  of  psychological  determinism.  The  initial  evi- 
dential ground  for  this  assumption  was  the  observation  that  apparently 
meaningless  hysterical  symptoms,  previously  attributed  to  a  somatic 

70  This  is  the  earliest  explicit  formulation  of  metapsychology;  see  Freud  [117, 
p.  114]. 


etiology,71  disappeared  when  the  patient,  in  hypnosis,  related  them  to  past 
experiences,  thoughts,  feelings,  or  fantasies,  and  thus  endowed  them  with 
meaning  and  psychological  "cause."72  This  success  at  tying  apparently 
arbitrary  pathological  behavior  into  a  causal  psychological  network 
served  as  the  empirical  point  of  departure  for  the  venture  into  the 
broader  realm  of  dreams  [98],  parapraxes  [99],  etc.  The  success  in 
"interpreting"  these  resulted  in  the  further  and  apparently  limitless 
generalization  of  this  assumption,  on  which  all  the  other  initial  assump- 
tions of  the  theory  rest.  Clearly,  the  empirical  evidence  alone,  without 
the  background  factors  discussed  in  the  introduction  of  this  essay,  might 
not  have  given  rise  to  the  assumption  of  psychological  determinism. 

2.  The  assumption  of  unconscious  psychological  processes.  The  ob- 
servation that  in  hypnosis  and  in  the  course  of  free  associating  patients 
become  aware  of  past  experiences,  or  of  relations  between  them,  or  of 
relations  between  past  and  present  experiences,  led  to  the  assumption 
of  the    "nonconscious"   survival   of  such   experiences  and   the    "non- 
conscious"  existence  of  such  relationships  [35,  95].  But  only  the  discovery 
that  such  nonconscious  experiences  and  relationships  are  subject  to  rules 
(e.g.,  the  pleasure  principle  and  the  mechanisms  of  the  primary  process) 
different  from  those  of  our  conscious  behavior  and  thinking  made  the 
above-mentioned  memory  phenomena  (already  observed  by  Charcot,  as 
well  as  Bernheim)  [see  Breuer  and  Freud,  35,  chap.  1]  into  evidence  for 
the  assumption  of  unconscious  psychological  processes  [98].  The  essence 
of  this  assumption  is  that  it  conceptualizes  these  observations  in  psy- 
chological terms,  though  the  processes  inferred  from  them  are  subject  to 
rules  different  from  those  of  the  familiar,  conscious  psychological  proc- 
esses. In  other  words,  it  refuses  to  treat  the  nonconscious  as  somatic  and 
the  nonlogical  as  nonpsychological.  It  rejects  both  consciousness  and 
logical  relations  as  necessary  criteria  of  psychological  processes,  and  thus 
arrives  at  the  concept  of  unconscious  psychological  processes  abiding  by 
rules  other  than  those  of  conscious  processes.  This  assumption  gained 
powerful  corroborative  evidence  from  the  study  of  dreams  [98,  p.  540]. 

3.  The  assumption  of  unconscious  psychological  forces  and  conflicts. 
The  evidence  for  unconscious  psychological  processes  did  not,  in  the 
beginning,    necessitate    the    assumption    of    unconscious    psychological 
forces  and  conflicts.  Breuer's  hypnoid  assumption  and  Freud's  trauma 

nEven  the  psychologically  minded  French  school,  Charcot,  Janet,  etc.  [see 
309],  subscribed  to  this. 

72  For  the  detailed  reports  of  these  observations,  see  [35].  The  theoretical  section 
of  that  volume  contains  a  fragmentary  and  simplified  version  of  the  neuropsy- 
chological  theory  Freud  developed  in  the  Project  [94,  appendix]  to  account  for 
these  observations.  These  two  theoretical  statements  are  the  predecessors  of  Freud's 
theory  contained  in  the  seventh  chapter  of  The  Interpretation  of  Dreams  [98]. 

The  Structure  of  Psychoanalytic  Theory  113 

and  retention  assumption  [35]  seemed  to  account  for  the  unconscious 
character  of  these  processes.  The  evidential  ground  for  the  assumption 
of  unconscious  psychological  forces  was  Freud's  discovery  that  much 
of  what  his  patients  reported  to  him  was  not,  as  he  initially  believed, 
unconscious  memories  of  actual  experiences,  but  rather  unconscious 
fantasies  [94,  p.  215,  letter  no.  69,  1897].  The  assumption  of  unconscious 
forces  was  to  account  for  the  agent  which  creates  these  fantasies  and 
brings  them  to  the  patient's  consciousness  in  hypnosis  and  in  free  as- 
sociations, as  well  as  for  the  agent  which,  before  and  in  the  course  of 
therapy,  prevents  them  from  becoming  conscious.  Led  by  the  libidinal 
content  of  these  fantasies,  Freud  assumed  that  the  unconscious  force 
which  creates  them,  and  makes  them  conscious  in  the  course  of  therapy, 
is  the  sexual  drive.  In  turn,  he  conceived  of  the  forces  which  clash  with 
the  sexual  drives,  divert  them  into  symptoms,  and  block  the  path  to 
consciousness  of  the  fantasies  which  they  gave  rise  to,  as  those  of  the 
censorship :  the  ego  drives. 

Thus  the  initial  evidential  ground  for  this  assumption  comprised 
observations  pertaining  to  unconscious  fantasies,  to  their  becoming  con- 
scious in  therapy,  to  the  resistance  against  their  becoming  conscious,  and 
to  the  relation  between  these  fantasies  and  the  symptoms. 

4.  The  assumption  of  psychological  energies  and  their  drive  origin. 
The  observation  that  recall  of  traumatic  experiences,  when  accompanied 
by  affect,  results  at  times  in  the  disappearance  of  symptoms  and  anxiety, 
and  at  other  times  in  their  replacement  by  other  symptoms  and  anxiety 
equivalents,  suggested  that  a  displaceable  and  transformable  quantity 
was  involved  in  the  psychological  processes  underlying  symptom  forma- 
tion. Before  he  developed  the  concept  of  unconscious  forces,  Freud  as- 
sumed that  this  quantity  was  the  affect,  which  when  not  expressed  (i.e., 
"dammed  up53)  was  either  transformed  into  anxiety  or  displaced  into  a 
somatic  organ  (conversion)  or  a  thought  process  (e.g.,  obsession).  After 
he  developed  the  concept  of  drives,  this  quantity  was  conceived  of  as 
drive  energy  (cathexis). 

The  force  concept  alone  could  not  account  for  the  observation  that 
blocking  a  drive  action  results  in  behavior  different  in  direction  and 
form  from  that  expected  of  the  drive;  this  observation  became  the 
evidential  ground  for  the  assumption  of  psychological  energies  and  of  a 
conservation  principle  pertaining  to  them.  These  psychological  energies, 
analogous  to  those  of  physics,  being  nondirectional  (scalar)  could, 
through  their  displacements  and  transformations,  account  for  the  "work" 
performed  by  the  psychological  force  in  forms  unlike,  and  at  points  not 
coinciding  with,  that  expected  of  them.  This  assumption  when  combined 
with  that  of  the  instinctual  origin  of  the  unconscious  psychological  forces 
led  to  the  assumption  of  the  drive  origin  of  psychological  energies. 


The  evidence  for  the  assumption  of  an  entropy  principle  and  a 
principle  of  least  action  lay  in  the  observation  of  the  difference  between 
those  two  kinds  of  behavior  characteristics  which  were  conceptualized  as 
manifestations  of  the  primary  and  secondary  processes.  The  prevalence 
of  the  first  kind  of  characteristic  makes  a  behavior  peremptory  and  over- 
valent,  tolerating  neither  delay  nor  detour,  as  though  it  were  striving 
for  immediate  discharge  of  a  great  quantity  of  excitation.  These  char- 
acteristics of  obsessional  and  delusional  ideas,  compulsive  rituals,  hys- 
terical tantrums,  etc.,  served  as  the  evidential  ground  for  the  assumption 
of  an  entropy  (pleasure-pain)  principle.  The  prevalence  of  the  second 
kind  of  characteristic  makes  a  behavior  amenable  to  interruption,  delay, 
and  detour,  as  though  it  involved  potentials  without  significant  intensities. 
These  characteristics  of  goal-directed  action,  and  of  ordered  logical 
thought,  were  the  initial  evidence  for  the  assumption  of  a  principle 
of  least  action. 

It  may  be  objected  that  this  discussion  has  not  dealt  with  the  as- 
sumption of  the  ultimate  determination  of  all  behavior  by  unconscious 
drives,  which,  with  its  emphasis  on  ultimate  and  all,  is  indeed  one  of  the 
assumptions  of  early  psychoanalytic  theory.  We  bypass  it  here  because  it 
is  actually  a  combination  of  the  assumptions  we  have  discussed. 

B.  Strategic  Choice  of  Initial  Evidential  Grounds 

The  question  why  the  observations  which  served  as  the  initial  evi- 
dential ground  for  the  assumptions  of  the  system  were  considered 
strategic  is  in  a  sense  irrelevant  to  the  theory  of  psychoanalysis.  The  initial 
situation  was  not  that  the  phenomena  of  pathology  were  considered 
strategic:  they  were  the  material  which  posed  the  problems  to  be  coped 
with.  The  theory  grew  up  on  the  soil  of  the  neuroses,  their  pathology 
and  their  therapy.  It  was  from  there  that  it  branched  out  into  a  relentless 
and  ever  more  diversified  endeavor  to  show  that  its  system  of  conceptual 
relationships,  though  it  was  designed  to  explain  pathological  (apparently 
arbitrary  and  psychologically  meaningless)  phenomena,  can  also  give  an 
adequate  causal  account  of  the  obviously  meaningful  phenomena  of 
normal  psychological  life. 

Pathology  was  (as  Virchow  recognized  in  biology)  strategic  for  the 
study  of  normal  processes.73  It  showed  that  the  so-called  normal  state  of 

73 Pathology  had  still  another  role  in  the  development  of  psychoanalysis:  ex- 
cept for  love  and  mortal  fear,  only  actual  suffering  and  the  hope  of  relief  could 
have  prompted  a  man  to  permit  another  that  relatively  unlimited  access  to  his 
privacy  which  opened  the  door  for  psychoanalysis  to  the  exploration  of  its  initial 
evidential  ground.  That  this  opening  is  at  the  same  time  an  obstacle  to  the  inde- 
pendent verification  of  psychoanalytic  theories  is  as  -natural  as  it  appears  para- 
doxical at  first  sight. 

The  Structure  of  Psychoanalytic  Theory  115 

affairs  which  we  take  for  granted  is  only  one  of  many  possibilities.  Thus 
it  opened  the  road  to  causal  analysis,  by  means  of  which  psychoanalysis 
consistently  and  successfully  shattered  the  barriers  between  the  normal 
and  the  pathological,  the  infantile  and  the  adult,  the  recondite  and  the 
obvious,  the  exceptional  and  the  commonplace  [98,  pp.  538-540].  It  is 
not  a  historical  accident  that  Freud's  theory  grew  out  of  the  study  of 

Pathology  and  its  therapy  were  strategic  for  the  discovery  of  the 
commonalities  of  normality  and  pathology,  but  they  proved  less  strategic 
for  the  discovery  of  the  differences  between  them.  Only  slowly,  with  the 
development  of  psychoanalytic  ego  psychology,  has  psychoanalysis  begun 
to  rediscover  the  differences  between  the  pathological  and  the  normal, 
the  infantile  and  the  adult,  the  maladaptive  and  the  adaptive.  So  far, 
the  concepts  of  structure,  autonomy,  adaptation,  and  reality  are  the  main 
tools  the  theory  uses  in  its  endeavor  to  discover  these  differences.  These 
are  the  very  concepts  which  distinguish  psychoanalysis  from  the  genetic 
reductionist  theories  which  see  no  cleavage  between  the  normal  and  the 
pathological,  the  adult  and  the  infantile,  as  well  as  from  G.  AUport's 
[8]  and  kindred  theories  which  see  a  sharp  cleavage  between  them. 

We  cannot  leave  this  discussion  without  dwelling,  at  least  briefly, 
on  the  methods  by  which  the  initial  evidence  was  obtained. 

Nowadays  methodology  is  in  vogue,  and  all  considerations  of  method 
and  technique  are  dignified  by  that  name.  Yet  one  essential  methodo- 
logical task — the  study  of  the  relationship  between  a  theory  and  the 
method  of  observation  by  which  the  data  it  explains  are  obtained — 
is  rarely  pursued.  The  question  is:  to  what  extent  does  a  theory,  based 
on  data  obtained  by  a  given  method,  reflect  the  nature  of  the  data  itself, 
and  to  what  extent  does  it  reflect  the  method  of  data-gathering  and  its 
limitations?  The  man  who  shouts  into  an  empty  room  is  likely  to  hear  his 
own  echo;  likewise  the  investigator  may  get  back  little  more  than  what 
he  has  already  built  into  his  method.  For  instance,  we  need  to  know  to 
what  extent  the  "laws  of  learning"  are  laws  of  human  nature,  and  to 
what  extent  they  are  artifacts  of  the  method  used  by  associationists  and 
conditioners  to  "establish"  them.  Likewise  to  what  extent  does  psy- 
choanalytic theory  reflect  human  nature,  and  to  what  extent  does  it 
reflect  Freud's  methods  for  studying  human  nature? 

Methodological  study  is  likely  to  reveal  that  some  psychoanalytic 
methods  (for  instance,  the  therapist-patient  two-group)  had  a  defining 
influence  on  psychoanalytic  theory  [see  Rapaport,  260].  Although  we  can- 
not pursue  this  problem  further,  we  want  to  suggest  that  methodological 
analysis  may  well  lead  to  a  distinction  between  a  general  psychoanalytic 
theory  which  is  little  dependent  on  these  methods,  and  a  specific  psycho- 
analytic theory  which  is  greatly  dependent  on  them.  In  contrast  to  the 


specific  theory,  the  general  theory  should  be  readily  testable  by  methods 
other  than  those  by  which  the  initial  evidence  for  it  was  obtained.  In  this 
essay  we  have  centered  on  those  aspects  of  the  theory  which  are  not 
obviously  dependent  on  these  methods,  and  have  tried  to  avoid  the 
concepts  which  obviously  are  tied  to  them,  like  transference,  interpreta- 
tion, etc.74 

C.  The  Relation  of  the  Observations  to  the  Theory 

In  his  outline,  Dr.  Koch  asks  us  to  select  the  chief  empirical  in- 
dependent and  dependent  variables  of  the  theory  and  to  demonstrate 
their  linkage  to  its  systematic  independent  and  dependent  variables.  In 
Section  II.  K.,  we  discussed  the  difficulties  in  the  way  of  such  an  at- 
tempt. To  minimize  repetition,  we  will  illustrate75  the  relation  of  an 
empirical  observation  to  the  variables  of  the  theory. 

Let  us  take  the  words  of  a  man  who  utters  the  sentence,  "Now  things 
are  becoming  queer,53  and  let  us  provide  the  context  from  which  psy- 
choanalytic theory  will  adduce  its  explanation  of  this  verbal  behavior: 

This  behavior  occurred  in  the  course  of  a  discussion  in  a  group. — The 
other  members  responded  to  it  with  consternation. — The  man  was  be- 
wildered by  this  response. — Later  he  found  out  that  he  had  said  "queer" 
and  not  "clear"  as  he  had  intended  to  do,  and  as  he  thought  he  had 
done. — He  was  embarrassed  by  this  discovery. — The  discussion  concerned 
a  mismanagement  of  the  group's  affairs. — The  subject's  utterance  fol- 
lowed an  explanation  by  the  chairman  of  the  group. — The  chairman  at- 
tributed the  mismanagement  to  a  misunderstanding  by  the  treasurer  of 
an  instruction  given  by  him,  and  not  to  any  malicious  intent. — The  chair- 
man commanded  the  unquestioned  respect  of  the  group  and  also  wielded 
considerable  power  otherwise. 

In  terms  of  common-sense  psychology,  we  are  dealing  here  with  a 
slip  of  the  tongue. 

In  descriptive  terms:  the  subject's  conscious  intention  was  to  agree 
with  the  chairman's  explanation.  He  did  not  carry  out  this  intention,  but 

74 An  example  to  highlight  the  relationship  of  method  and  theory:  it  appears 
that  H.  S.  Sullivan  [314],  taking  as  his  point  of  departure  the  psychoanalytic  meth- 
ods of  the  two-group  and  participant  observation,  arrived  at  a  theory  of  per- 
sonality which  dissolves  the  concept  of  the  individual  and  conceives  of  the  person 
as  one  of  the  quasi-stable  foci  in  the  network  of  interpersonal  relationships.  In 
Sullivan's  theory  then,  the  method  of  investigation  and  the  transference  concept 
based  on  it  came  to  play  a  dominant  role,  with  the  consequence  that  the  theory 
overrides  a  crucial  characteristic  of  the  nature  of  the  subject  matter,  namely,  the 
individuality  of  the  person.  Individuality  to  Sullivan  appears  as  a  noxious  anti- 
scientific  myth,  which  he  reduces  to  the  personification  function  of  the  self-system. 

's  This  example  simplifies  an  actual  situation  by  eliminating  obscure  points,  to 
avoid  lengthy  explanations  of  peripheral  matters. 

The  Structure  of  Psychoanalytic  Theory  117 

instead  expressed  himself  in  a  way  that  caused  consternation;  he  was 
unaware  both  of  not  having  carried  out  his  intention,  and  of  his  con- 
sternation-arousing utterance.  When  he  was  told  what  he  had  said,  he 
became  embarrassed. 

In  terms  of  a  data  language:  the  independent  variable  (conscious 
intention)  determined  a  value  of  the  dependent  variable  (verbal 
utterance  of  agreement  and  conscious  awareness  of  it).  However,  an 
intervening  variable  determined  another  value  of  the  action  component 
of  the  dependent  variable  (dissenting,  consternation-arousing  verbal 
utterance).  The  intervening  variable  left  the  conscious-awareness  com- 
ponent of  the  dependent  variable  unaltered.  A  second  intervening  vari- 
able (external  reality:  information)  altered  the  conscious-awareness  com- 
ponent of  the  dependent  variable  and  determined  its  affective  aspect 
(embarrassment) . 

This  formulation  is  not  "neutral":  it  implies  that  the  independent 
variable  is  a  "motivation"  (intention).  Indeed,  even  the  descriptive 
terms  imply  this.  Before  Freud,  at  least  the  common-sense  term  "slip  of 
the  tongue"  was  neutral,  but  it  is  not  neutral  now.  Let  us  attempt  a 
crude  associationist  formulation,  to  show  that  data  languages  are  in- 
separable from  construct  languages  and  thus  cannot  be  neutral:  the 
chairman's  explanation  was  associated  in  the  subject  to  an  approving 
verbal  statement;  the  subject's  actual  response,  however,  was  linked  to 
the  chairman's  explanation  by  stronger  associative  bonds;  the  clash  of  the 
two  associative  complexes  resulted  in  a  compromise  in  which  one  of  the 
complexes  determined  the  awareness,  while  the  other  determined  the 
verbal  response  of  the  subject. 

In  terms  of  psychoanalytic  construct  language:  the  subject's  con- 
scious intention  is  referred  to  a  socially  adaptive  ego  interest.  The  failure 
to  carry  out  the  intention  is  referred  to  an  id  motivation.  The  unaware- 
ness  of  the  failure  is  referred  to  an  unconscious  ego  motivation  conflicting 
with  this  id  motivation.  The  unawareness  of  the  actual  verbal  expression 
used  is  referred  both  to  the  unconscious  (id)  motivation  which  was  ex- 
pressed, and  to  the  unconscious  ego  controls  (defenses)  which,  though 
they  failed  to  prevent  the  use  of  the  ego's  executive  apparatus  by  the  id 
motivation,  succeeded  in  preventing  its  access  to  consciousness  (com- 
promise). The  acute  embarrassment  is  referred  to  the  affect  manifesta- 
tion of  the  clash  between  the  unconscious  motivations  and  the  restored 
ego  control. 

Let  us  take  a  closer  look  at  the  concepts  involved.  The  unawareness 
is  obviously  the  referent  of  the  descriptive  concept  unconscious.  It  is 
likewise  obvious  that  the  intent  to  say  "clear"  is  a  conscious  motive. 
But  it  is  an  inference  that  this  motive  is  a  force  and  it  is  a  further  in- 
ference that  saying  "queer"  indicates  the  presence  of  another,  un- 


conscious,  force.  It  is  a  still  further  inference  that  a  third  force  is  also 
involved  which  prevents  conscious  awareness  both  of  the  failure  of  the 
conscious  intention  and  the  success  of  the  unconscious  intention.  It  is  yet 
a  further  inference  that  the  latter  two  forces  conflicted  and  reached  a 
compromise,  indicated  both  by  the  "clang"  similarity  of  the  words 
"clear35  and  "queer/5  and  by  the  fact  that  the  unconscious  force  attained 
control  of  the  executive  apparatus  but  did  not  gain  access  to  conscious- 
ness. Thus  we  see  that  some  of  the  concepts  involved  here  are  close  to  the 
observations,  wrhile  others  are  at  increasing  distance  from  them. 

In  clinical  inferences,  the  distance  between  observations  and  con- 
cepts may  seem  even  greater.  The  clinician  may  infer,  for  instance,  that 
the  slip  pertains  to  the  ancient  triangle  formed  by  the  subject,  his  older 
brother,  and  father,  which  was  reactivated  by  the  triangular  situation 
of  the  subject,  treasurer,  and  chairman.  He  may  even  go  further  and 
infer  that  homoerotic  and  aggressive  drives  involved  in  jealousy  are  the 
unconscious  forces  which  conflict  here  with  the  ego's  defenses  against 
them  and  interfere  with  ego  interests. 

No  wonder  psychologists  gained  the  impression  that  the  relation  of 
psychoanalytic  concepts  and  theories  to  observations  is  distant  and 
arbitrary.  But  is  this  impression  accurate?  Let  us  suppose  that  our  sub- 
ject volunteers  for  a  free  association  session,  and  his  associations  cluster 
around  the  treasurer  and  the  chairman,  rather  than  around  the  interests 
of  the  group.  Will  we  then  be  justified  in  inferring  that  the  agent  behind 
the  word  "queer"  is  an  unconscious  force  directed  toward  the  treasurer 
and  the  chairman?  Let  us  suppose  further  that  the  subject's  associations 
not  only  corroborate  that  this  unconscious  force  is  an  aggressive  drive, 
but  identify  it  as  being  of  a  jealous-suspicious  variety.  Let  us  finally  as- 
sume that,  in  the  course  of  these  associations,  the  subject  comes  to 
realize  that  he  actually  has  had  filial  feelings  toward  the  chairman  and 
vague,  poorly  understood  feelings  of  irritation  with  the  treasurer,  akin 
to  those  he  used  to  feel  toward  his  brother,  and  thereby  he  specifies  that 
the  unconscious  force  pertains  to  the  subject-brother-father  triangle. 

True,  in  this  sequence  the  concepts  (unconscious,  unconscious  im- 
pulse, unconscious  hostile  impulse,  unconscious  hostile  and  libidinal  im- 
pulse, unconscious  hostile  and  libidinal  infantile  impulse)  are  in- 
creasingly remote  from  the  slip  of  the  tongue  which  is  the  original 
observation.  But  the  associations,  too,  are  observations  and  the  in- 
creasingly remote  concepts  are  introduced  in  reference  to  these  additional 
observations.  Thus,  the  distance  between  observations  and  concepts  is 
not  as  great  as  it  seems  on  first  sight.  But  there  still  remains  a  difficulty : 
the  relationship  of  each  more  remote  concept  to  the  corresponding  ad- 
ditional observation  presupposes  the  less  remote  concepts.  For  instance, 
without  assuming  that  the  unawareness  of  the  subject  is  a  referent  of  the 

The  Structure  of  Psychoanalytic  Theory  119 

descriptive  concept  unconscious  and  without  assuming  the  unconsciously 
motivated  character  of  the  slip,  it  would  make  no  sense  to  infer  that 
the  subject's  associations  specify  the  pertinence  of  the  aggressive  impulse 
(indicated  by  the  slip)  to  the  treasurer  and  the  chairman. 

This  relationship  between  observations  and  concepts  is  common  to 
all  sciences:  observations  demonstrate  theoretical  relationships  only  to 
those  who  already  conceive  of  the  observed  in  terms  of  the  theory's  con- 
cepts. But  the  psychologist  seems  to  overlook  this  truism  when  it  comes 
to  psychoanalysis.  This  oversight  is  so  common  that  the  lack  of  systematic 
treatments  of  the  theory  alone  cannot  account  for  it.  There  must  be 
other  reasons,  and  a  few  of  these  will  be  conjectured: 

The  psychologist  is  accustomed  to  explicit — and,  indeed,  operational 
• — definitions  of  concepts  and  is  wary  of  psychoanalysis3  definitions  of 
concepts.  He  suspects  that  the  mutual  implications  of  its  concepts  hide 
a  vicious  circle.  In  the  lack  of  a  systematic  statement  of  the  theory,  we 
can  sympathize  with  his  wariness,  but  we  must  keep  in  mind  that  in 
physics  nobody  would  think  of  asking  for  an  explicit  definition  of  energy 
that  did  not  involve  the  concept  of  work  (which  in  turn  involves  the 
concepts  of  path  and  force,  which  in  turn  involve  mass  and  acceleration, 
which  in  turn  involve  time  and  velocity,  which  in  turn  involve  space 
and  time).  We  shall  not  dwell  here  on  the  demand  for  operational 
definitions:  Feigl  [71]  and  Frenkel-Brunswik  [92]  have  demonstrated 
that  in  this  matter  psychologists  have  tried  to  be  more  Catholic  than  the 
Pope,  and  that  operational  definitions  of  all  its  concepts  have  never  been 
demanded  of  any  science. 

The  clinical  psychoanalyst  is  deft  and  nonchalant  in  using  concepts 
at  a  great  distance  from  the  observations.  For  instance,  he  may  con- 
jecture from  the  word  "queer"  what  might  be  involved  in  this  slip,  by- 
passing the  intervening  observations  (e.g.,  associations)  and  concepts.  It 
may  be  a  well-supported  conjecture,  if  the  patient's  previous  productions 
converge  on  it;  or  it  may  be  a  poorly  supported  one,  if  the  analyst  is 
more  imaginative  than  careful.  It  may  even  help  the  patient  to  insight 
if  it  is  conveyed  to  him.  But  a  conjecture  it  remains  until  the  patient's  as- 
sociations or  other  productions  confirm  it.  Some  such  conjectures  are 
supported  by  so  much  experience,  and  pertain  to  relationships  so 
common,  that  they  are  almost  certain.  These  are  particularly  prone  to 
turn  into  cliches,  to  give  the  outsider  the  impression  of  arbitrariness  or 
of  an  uncanny  "second  sight,"  and  to  oversimplify  the  complexity  of 
the  theoretical  relationships  even  in  the  psychoanalyst's  mind.  Actually 
the  psychoanalyst's  use  of  these  may  not  differ  from  an  electrician's  use 
of  technical  terms  and  repair-  or  construction-procedures  without  his 
referring  to  or  even  being  aware  of  their  theoretical  implications.  When 
the  rules  of  thumb  of  clinical  psychoanalysis  are  equated  with  the  theory 


of  psychoanalysis,  the  observations  and  concepts  which  bridge  the  gap 
between  the  basic  concepts  and  the  initial  observations  are  inevitably 

The  psychoanalytic  writer  and  practitioner  is  inclined  to  speak  of 
psychoanalytic  concepts  and  theories  in  terms  of  "content."  The  con- 
tent of  the  word  "queer"  may  serve  as  an  example.  The  subject's  use 
of  this  word  is  conducive  (or  seductive)  to  the  conclusion  that  a  homo- 
sexual impulse  may  be  involved  in  the  production  of  this  slip.  The 
content  of  any  slip  may  suggest  the  nature  of  the  unconscious  motivation 
involved  in  it.  Content  is  an  important  guide  to  the  practitioner.  Many 
sensitive  and  experienced  psychoanalysts  are  to  a  great  extent  guided  by 
the  content  of  communications.  Others  are  guided  by  the  tone  of  voice 
or  other  expressions  of  emotion.  The  majority  of  the  contributions  to 
the  literature  tend  to  dwell  on  content  to  the  neglect  of  other  guides. 
What  is  lost  sight  of — and  the  practitioner  need  not  necessarily  keep  this 
in  focus  or  even  in  sight,  but  those  interested  in  the  theory  must — is  the 
functional  (and  thus  also  conceptual)  relationship  to  which  the  content 
is  a  guide.  The  word  "queer"  is — by  the  circumstances  of  its  utterance — 
a  compromise  formation  between  id  motivations  and  ego  controls.  This 
is  one  of  the  functional  relations  involved  in  this  slip.  This  slip's  content 
suggests  some  of  the  (aggressive  and  homoerotic)  impulses  involved  in 
this  functional  relation.  But  these  suggestions  make  theoretical  sense  only 
if  the  word  "queer"  and  the  circumstances  of  its  utterance  are  assumed 
to  be  referents  of  the  descriptive  concept  of  the  unconscious,  of  the  con- 
cept of  unconscious  motivation,  of  unconscious  drive  motivation,  of  un- 
conscious conflict,  and  of  resolution  by  compromise.  No  content  yields  its 
full  meaning  unless  its  formal  characteristics,  and  those  of  the  time,  locus, 
and  context  of  its  appearance,  are  taken  into  consideration,  that  is  to  say, 
abstracted.  The  content  of  the  word  "queer,"  and  of  any  communication, 
is  predictive  only  in  so  far  as  it  belongs  to  the  network  of  the  ideas  which 
represent  the  drives  involved.  In  other  words,  the  content  is  a  guide  be- 
cause it  belongs  to  a  network  of  formal  relationships.  Whether  it  is  the 
Oedipus  complex,  or  the  castration  complex,  or  an  anal  fixation,  or  a 
homosexual  impulse  to  which  the  content  refers,  it  does  so  always  by 
virtue  of  a  formal,  conceptual  relationship.  The  stress  on  content  seems 
to  be  one  of  the  main  causes  for  overlooking  the  relationship  between 
concepts  and  observables.  Psychoanalysts  are  not  the  only  ones  who 
make  a  direct  jump  from  content  to  unconscious  motivation :  Rorschach 
testers  and  other  projective  test  "experts"  do  it  too,  often  with  less  ex- 
perience and  always  with  less  collateral  information  to  go  by.76  Recently 
McClelland  [226]  called  on  psychologists  to  revive  their  interest  in 

76  See,   however,   Schafer's  [298]   treatment  of  content  as  a  guide  to   formal 

The  Structure  of  Psychoanalytic  Theory  121 

content.  McClelland  is  right,  it  is  high  time  to  begin  the  serious  experi- 
mental study  of  contents.  But  a  warning  of  the  pitfalls  is  in  place.  The 
more  familiar  the  contents  dealt  with  by  psychoanalytic  theory  become, 
the  greater  the  temptation  to  bypass  and  to  becloud  the  conceptual 


Psychoanalysis,  as  a  theory,  did  not  make  a  formal  study  of  the 
construction  of  its  functional  relationships.  Thus  whatever  can  be  said 
about  these  must  be  inferred.  The  preceding  sections  covered  this  ground 
as  much  as  seemed  feasible.  Here  we  can  add  only  a  discussion  of  the 
theory's  "function  form"  in  relation  to  the  Lewinian  and  S-R  function 

Lewin's  [219]  basic  function  form  is  B  =  f(P,E):  behavior  is  a 
function  both  of  the  environment  and  the  person.  Here  B  represents  be- 
havior at  large  and  not  any  specified  aspect  of  it;  E  represents  the  en- 
vironment as  the  person's  life  space  at  large  and  not  any  specified  part 
of  it;  and  P  represents  the  structural  and  tensional  characteristics  of  the 
intrapersonal  regions  at  the  time  of  the  behavior,  rather  than  the  person 
as  a  changing  historical  entity. 

B  =  f(P,E]  can  be  made  to  take  on  extreme  values  which  transform 
it  into  the  basic  S-R  function  form,  if  we  make  three  assumptions :  First, 
there  are  environmental  situations  in  which  behavior  (response)  is  in- 
variant in  respect  to  individual  differences  and  intrapersonal  changes; 
then  for  all  the  E  values  for  which  this  assumption  holds,  the  function 
changes  into  5  =  /(£).  This  situation  is  one  in  which  the  causal  texture 
of  the  environment  has  become  compelling.  In  terms  of  psychoanalytic 
ego  psychology,  this  is  a  situation  in  which  no  autonomy  from  external 
reality  obtains.  Second,  the  E  (life  space)  does  not  vary  from  subject  to 
subject  (such  variations  are  not  accounted  for  directly  by  this  equation). 
Third,  certain  behaviors  are  determined  not  by  the  E  in  general  but  by 

7T  Freud  seems  to  have  expressed  this  as  follows,  responding  in  a  letter  to 
Abraham  on  the  latter3  s  comments  on  "Mourning  and  Melancholia'9  [120]:  "... 
you  do  not  emphasize  enough  the  essential  part  of  my  hypothesis,  i.e.,  the  topo- 
graphical consideration  in  it,  the  regression  of  the  libido  and  the  abandoning  of 
the  unconscious  cathexis,  and  that  instead  you  put  sadism  and  anal-erotism  in  the 
foreground  as  the  final  explanation.  Although  you  are  correct  in  that,  you  pass  by 
the  real  explanation.  Anal-erotism,  castration  complexes,  etc.  are  ubiquitous  sources 
of  excitation  which  must  have  their  share  in  every  clinical  picture.  One  time  this 
is  made  from  them,  another  time  that.  Naturally  we  have  the  task  of  ascertaining 
what  is  made  from  them,  but  the  explanation  of  the  disorder  can  only  be  found 
in  the  mechanism — considered  dynamically,  topographically  and  economically" 
[193,  vol.  2,  p.  329]. 


a  specific  element  of  it;  that  is  to  say,  B  is  invariant  in  respect  to  E  ex- 
cept for  its  single,  specified  element  S.  If  we  then  use  the  term  re- 
sponse— R — for  the  so  specified  behavior,  and  the  term  stimulus — S — for 
the  so  specified  elements  of  E,  we  arrive  at  the  equation  R  =  f(S). 

Similarly,  B  =  f(PyE)  can  be  transformed  into  the  function  form 
of  that  phase  of  psychoanalysis  in  which  the  role  of  reality  was  negligible, 
and  the  role  of  the  person's  drives  and  defenses  paramount,  in  determin- 
ing behavior.  Embarking  on  such  a  transformation  we  must  note  that 
Lewin,  in  failing  to  establish  in  principle  the  relations  between  "geo- 
graphic" and  "hodological"  (life)  space,  left  the  door  wide  open  for  all 
those  who  wish  to  transform  E  into  a  function  of  P  and  thus  to  turn  the 
B  =  i(P,E]  equation  into  B  —  f(P)Js  Werner  and  Wapner's  "sensory- 
tonic"  theory  [323] — by  postulating  a  tonic  factor  in  perception — intro- 
duces P  into  the  perceived  E,  and  thus  also  leaves  an  opening  (however 
narrow)  for  such  a  transformation.  Others,  particularly  Brunswik, 
Heider,  Gibson,  and  recently  Klein  [197],  explicity  refused  to  do  so. 
x\ccording  to  Piaget's  studies,  the  "construction  of  reality"  is  an  onto- 
genetic  achievement  and  not  a  process  of  "imitative  learning"  or  "con- 
ditioning." Thus  in  his  theory,  too,  genetic  considerations  can  always 
resolve  E  into  E  ~g(P})  though  here  all  such  g  functions  are  actually 
of  the  form  En  =  gn(En-i,P},  and  P  itself  is  subject  to  historical  change 
[P»  =  An(P«-i3£n-i)].  Nevertheless,  every  genetic  theory  tends  to  trans- 
form B  —  f(P,E)  into  B  =  /(P) :  in  genetic  theories,  genetic  reduction- 
ism  is  always  a  temptation.79 

Can  any  dynamic  psychology  escape  such  a  reduction?  Allport's 
personalistic  psychology  has  perhaps  the  most  explicit  safeguards  against 
such  a  reduction,  which  turns  man  either  into  a  mechanism  ultimately  at 
the  mercy  of  its  environment,  or  into  a  solipsistic  creature  ultimately  at 
the  mercy  of  his  drives.  Allport's  [6,  8]  safeguard  against  both  these 
alternatives  is  expressed  in  his  concept  of  functional  autonomy,  which 
implies  that  whatever  the  genetic  (maturational  or  learning)  history  of 
a  function,  it  may  attain  autonomy  so  that  it  can  serve  as  an  irreducible 
basis  of  behavior.  In  psychoanalytic  theory,  a  similar  solution  was  in- 
dependently reached  by  Hartmann  and  expressed  in  his  concepts  of  the 
conflict-free  sphere  and  autonomous  ego  functions.  However,  it  should 
be  noted  that  his  is  a  concept  of  relative  autonomy :  functions  and  struc- 
tures have  only  limited  autonomy  from  the  drive  or  learning  process 
from  which  they  arose;  for  instance,  behaviors  determined  by  such  struc- 
tures may  be  over  determined  by  drives;  they  may  be  used  by  drives  as 

78  Brunswik  and  Heider  point  out  that  Lewin's  environment  is  "encapsulated.55 
G.  S.  Klein  points  out  that  Lewin  disregarded  the  "inherent  structure33  of  the  en- 
vironment and  centered  exclusively  on  its  perceived  structure. 

79  But  see  Erikson  [61,  62,  66]. 

The  Structure  of  Psychoanalytic  Theory  123 

means;  and  under  stress  their  autonomy  does  not  necessarily  hold  and 
they  may  yield  to  ontogenetically  earlier  forms.80  The  implications  of 
Piaget's  [254]  genetic  theory  of  intelligence  are  similar:  the  new  and 
higher-level  "circular  reactions15 — and  the  "schemata"  corresponding  to 
them — attain  independence  from  the  schemata  from  which  they  arose. 
But  the  lower-order  schemata  and  circular  reactions  are  not  replaced 
by  the  higher  ones  and  may  always  be  reactivated  when  the  latter  do  not 
provide  the  means  of  coping  with  the  situation  encountered. 

B  =  i(PJ£]  implies  that  no  broadly  valid  relationships  of  the 
B  =  f(P)  or  the  B  =  f(E]  type  are  possible.  If  the  B  =  f(P]  function 
form  is  to  attain  validity  for  more  than  a  narrowly  specified  range,  E 
must  be  introduced  into  it  as  an  intervening  variable.  The  same  consider- 
ation holds  for  the  validity  of  B  =  f(E),  and  demands  the  introduction 
of  P  as  an  intervening  variable.  But  neither  E  nor  P  is  a  simple  variable : 
both  are  functions  of  other  variables,  which  can  be  held  constant  only 
for  a  narrow  range  of  conditions.  If  we  take  it  for  granted  that  the 
person's  experience  of  his  life  space  is  our  basic  information  about  his 
geographic  environment,  then  £  is  a  function  of  the  geographic  environ- 
ment and  P;  and  if  P  changes — as  it  does — with  experience,  then  it  is 
a  function  of  preceding  P's  and  E's. 

Let  us  approach  the  problem  from  another  angle.  The  extreme  values 
which  make  B  =  f(P,E)  go  into  either  B  =  f(E]  or  B  =  /(P)  imply 
that  certain  one-to-one  relations  between  stimulus  and  response,  drive 
and  behavior  are  possible.  But  we  know  empirically  that  by  and  large 
this  is  an  untenable  assumption,  since  the  single  S  (stimulus)  is  hardly 
ever  the  only  effective  factor  in  E  and  the  single  D  (drive)  is  hardly 
ever  the  only  effective  factor  in  P.  So  when  S  or  D  is  chosen  as  the  in- 
dependent variable,  the  other  factors  come  into  play  as  intervening 
variables.  In  other  words,  the  remarkable  thing  about  human  behavior 
is  that  man  often  meets  diverse  stimuli  by  the  same  behavior,  and 
identical  stimuli  often  elicit  diverse  responses.  Likewise  with  motivations : 
the  same  motive  may  be  expressed  by  a  wide  variety  of  behaviors  or 
satisfied  by  a  variety  of  objects,  and  a  great  variety  of  motives  may  be  ex- 
pressed by  the  same  behavior  or  satisfied  by  the  same  object  [see  Frenkel- 
Brunswik,  91,  and  Gill,  146].  Therefore,  if  stimuli  or  motives  are  used 
as  independent  variables,  it  becomes  necessary  to  introduce  intervening 
variables  to  account  for  the  flux  of  the  dependent  variable.  Thus,  learn- 
ing theory  introduced  sets,  attitudes,  etc.,  as  intervening  variables,  to  save 
the  R  =  j(S]  function  form.  Where  P  at  large  is  the  independent 
variable,  E  will  serve  as  the  intervening  one,  and  vice  versa.  In  psy- 

80  This  is  only  a  possible  consequence  of  stress,  not  a  necessary  one.  Among 
others,  Jacobson  [189]  and  Bond  [30]  report  observations  of  increased  autonomy 
and  efficiency  under  stress.  See  also  [280]. 


chologies  where  the  P  is  differentiated,  and  the  motive  (e.g.,  the  drive) 
is  the  independent  variable,  structures  (defenses,  controls,  etc.)  will 
appear  as  intervening  variables.  In  those  psychologies  wrhere  the  E  is 
differentiated,  context  and  "setting"  will  appear  as  intervening  variables. 
In  conclusion:  by  the  very  nature  of  psychoanalytic  theory,  inter- 
vening variables  are  indispensable  in  its  function  forms.  It  would  seem 
that  this  holds  true  for  all  dynamic  psychologies,  and  the  range  of  inter- 
vening variables  they  will  use  will  depend  upon  the  degree  of  autonomy 
they  assign  to  the  structures  and  functions  involved  in  the  phenomena 


It  would  be  simplest  to  restate  at  this  point  that  psychoanalysis  as  a 
discipline  has  not  attempted  quantification,  and  avoid  the  whole  issue. 
One  could  express  justified  impatience  with  the  "furor  of  measuring" 
which  has  gripped  psychology,  partly  by  referring  to  the  history  of 
evolution  theory,  in  which  precious  little  was  measured  for  a  long  while, 
and  partly  by  referring  to  how  much  is  being  feverishly  measured  in 
psychological  laboratories  without  good  reason.  Sometimes  one  has  the 
impression  that  the  hope  in  such  measuring  is  well  expressed  in  the 
Hungarian  proverb,  "Even  the  blind  hen  does  at  times  peck  a  grain." 

But  the  issue  of  quantification  cannot  be  dismissed  lightly.  Psycho- 
analysis— like  all  other  sciences — orders,  equates,  compares,  and  dis- 
tinguishes observables,  and  these  procedures,  once  made  precise,  reveal 
themselves  as  mathematical  operations  [cf.  Piaget,  252,  vol.  1].  Thus  all 
sciences,  in  striving  to  make  their  assertions  precise,  move  toward  a 
mathematization  of  the  relationships  they  establish  by  their  procedures. 
Since  mathematization  may  be  either  metric  or  nonmetric,  quantifica- 
tion is  only  one  form  of  it.  In  contrast  to  the  customary  quantification, 

81  It  can  be  argued  that  the  use  of  intervening  variables  does  not  depend  on  the 
degree  of  autonomy  the  system  assigns  to  structures  and  functions,  but  rather  on 
the  observational  method  by  which  the  data  the  theory  accounts  for  are  obtained. 
For  example,  R.  R.  Holt  (personal  communication)  comments:  "Skinner  does  not 
need  intervening  variables  because  he  forces  behavior  into  a  narrow  compass 
where  S  and  R  can  be  directly  related  mathematically."  It  is  probable,  however, 
that  Skinner's  procedure,  too,  is  rooted  in  a  choice  of  autonomous  (automatized) 
relationships.  It  would  seem  that  the  degree  of  autonomy,  as  well  as  the  role 
assigned  to  intervening  variables,  depends  on  the  observational  method  used.  The 
psychoanalytic  method  alone  scarcely  allowed,  and  certainly  did  not  require,  the 
theory  to  introduce  the  concept  of  autonomy.  It  was  introduced  when  data  ob- 
tained by  other  observational  methods  were  also  considered  by  the  theory.  Con- 
versely, Hebb  [169]  seems  to  have  realized,  when  he  considered  methods  of  ob- 
servation (e.g.,  Senden's)  other  than  conditioning,  that  the  S-R  relation  is  not 
free  (autonomous)  from  what  passes  between  the  S  and  the  R. 

The  Structure  of  Psychoanalytic  Theory  125 

Lewin  [217]  attempted  to  introduce  topology  and  Piaget  [253]  to  intro- 
duce group  theory  into  psychology  as  nonmetric  mathematizations.  Non- 
metric  mathematization  does  not  necessarily  exclude  quantification;  in- 
deed it  is  not  clear  whether  Lewin's  quantifications  are  the  result  of  his 
success  or  failure  in  nonmetric  mathematization.  The  problem  of  metric 
mathematization  is  what  kind  of  quantification,  if  any,  is  appropriate  to 
what  kind  of  psychological  relationship. 

It  is  probable  that  psychoanalysis  has  not  developed  a  mathematiza- 
tion, and  that  academic  psychology  has  not  gotten  far  with  its  strenuous 
efforts  at  quantification,  because  they  are  both  sciences  in  an  early  phase 
of  development.  Whether  it  is  metric  or  nonmetric  mathematization  that 
psychoanalysis  is  headed  for  is  hard  even  to  discuss  at  this  point.  It  would 
certainly  be  premature  to  judge  that  quantification  is  the  kind  of  mathe- 
matization which  is  appropriate  to  psychoanalytic  theory.  The  following 
discussion  of  quantification  implies  no  such  judgment,  but  merely  this 
attitude:  since  the  question  of  quantification  has  been  raised,  and  since 
quantification  may  prove  to  be  the  mathematization  appropriate  to  psy- 
choanalysis, some  of  the  problems  it  involves  should  be  explored.  This 
section  will  dwell  on  two  topics:  on  the  quasi-quantitative  concept  of 
cathexis,  which  of  all  the  concepts  of  the  theory  seems  to  call  most 
urgently  for  quantification,  and  on  the  kind  of  quantification — if  any — 
required  by  the  theory. 

A.  Cathexes82 

The  psychoanalytic  theory  contains  quasi-quantitative  concepts.  The 
most  conspicuous  of  these  are  the  drives,  which  are  conceived  of  as 
forces,  and  the  cathexes  they  expend,  which  are  conceived  of  as  quantities 
of  energy.  Why  then  have  these  not  been  measured?  To  answer  this 
question  it  is  necessary  to  discuss  the  distinctions  psychoanalytic  theory 
makes  between  various  forms  of  energy. 

1.  The  muscular  energy  of  behavior  is  not  the  psychological  energy 
that  psychoanalytic  theory  speaks  of:  the  psychological  forces  which  in 
their  work  expend  psychological  energy  only  release  the  forces  that  ex- 
pend the  biochemical  energy  of  muscles. 

2.  Psychological  energy  (in  the  main)  is  considered  as  of  drive  origin, 
and  to  account  for  its  major  forms  of  manifestation,  two  transformation 
processes  are  postulated :  binding  and  neutralization.  Both  of  these  result 
in  forms  of  energy  (bound,  neutralized]  which  differ  from  the  original 
(mobile]  form  of  drive  energy. 

3.  These  three  forms  of  energy,  and  the  two  major  processes  of  trans- 
formation, may  be  characterized  as  follows : 

82  For  references,  see  pp.  91-93,  113-114,  above. 


a.  Mobile  energy  abides  by  the  pleasure  principle:  It  tends  toward 
Immediate  discharge  by  the  shortest  route,  and  brooks  neither  delay  nor 
detour.  These  direct  discharges  may  take  the  form  of  action,  idea,  or 
affect  (e.g.  impulse  action,  compulsive  ritual,  random  action;  delusion, 
obsessional  idea;  and  uncontrolled  affect  storm,  such  as  a  tantrum  or 
panic,    etc.).    The    processes   which   expend    mobile    energy    are    con- 
ceptualized as  primary  processes.  They  make  use  of  several  mechanisms 
(i.e.,  specific  transformations  of  energy)  as,  for  instance,  condensation, 
displacement,  substitution,  symbolization,  etc.  These  mechanisms  come 
into  play  in  all  facets  of  behavior,  though  they  are  commonly  illustrated 
by  examples  taken  from  ideation  (e.g.,  dreams).  Mobile  energy,  though 
its  function  is  particularly  well  illustrated  by  the  behaviors  mentioned 
above,  does  not  refer  to  a  class  of  behaviors  but  to  a  component  of  all 

b.  Bound  energy  is  defined  as  energy  tied  up  in  structures.  Breuer 
[35,  pp.  140-141]  compared  it  with  the  tonic  innervation  of  muscles. 
The  structures,  the  building  of  which  amounts  to  a  binding  of  energy, 
are  conceptualized  on  the  one  hand  as  those  controlling  and  defensive 
structures  of  the  ego  which  make  ordered  thought  as  well  as  controlled 
affect  and  goal-directed  behavior  possible,  and  on  the  other  hand  as  those 
which  are  the  means  (information;  habits;  concepts;  anticipatory,  gram- 
matical, syntactic,  and  logical  patterns,  etc.)  used  by  ordered  thought, 
controlled  affect,  and  goal-directed  behavior.  The  processes  made  pos- 
sible by  these  defensive  and  controlling  structures,  and  by  these  structures 
of  means  character,  are  conceptualized  as  secondary  processes.  The  trans- 
formation of  binding  changes  mobile  energies  into  bound  energies.  The 
structures  thus  created  counteract  the  mobility  of  unbound  energies,  and 
also  serve  as  the  means  (apparatuses)  by  which  the  latter  are  expended 
and  controlled.  Compared  with  the  great  energy  expenditure  in  primary 
processes,  the  structures  formed  by  binding  can  function  (autonomously) 
with  a  minimal  expenditure  of  psychological  energy,  and  by  controlling 
the   discharge   of  mobile    (great  intensity)    energies   they   create   high 
potentials  for  action.  Like  physical  mechanisms,  they  transform,  save, 
and  expend  energy.  The  concepts  of  the  binding  process  and  of  the 
structures  which  it  creates  account  for  that  aspect  of  the  psychological 
organization  which  does  not  reduce,  but  maintains  or  even  increases, 
tension  [see  Freud,  98,  pp.  533-534;  and  Allport,  8]. 

£.  Neutralized  energy  is  defined  as  energy  whose  tendency  to  follow 
the  pleasure  principle  (direct  immediate  discharge)  is  decreased.  This 
definition  implies  a  spectrum  of  energy  forms,  ranging  from  barely 
neutralized  to  highly  neutralized  energies.  The  process  of  neutralization 
is  defined  as  the  transformation  by  which  drive  energies,  the  ideal  type 
of  which  is  considered  nonneutralized,  are  transformed  into  energies  of 

The  Structure  of  Psychoanalytic  Theory  127 

various  degrees  of  neutralization.  The  discharge  (entropic)  tendency  is 
common  to  all  energy  (physical  as  well  as  psychological) :  how  can  we 
conceive  of  energies  which  abide  by  it  only  more  or  less?  The  limitations 
of  our  systematic  knowledge  permit  only  an  answer  by  analogy: 

The  entropy  principle  certainly  obtains  for  closed  systems  of  physical 
energy,  but  organisms,  like  other  open  systems  and  like  man-made 
mechanical  structures,  postpone  and  obstruct  the  operation  of  the 
entropic  tendency.  Organisms  do  this  by  being  structured  and  by  build- 
ing further  structures  [cf.  Schroedinger,  305].  Man-made  structures  do  it 
by  preventing  expenditures  of  kinetic  energy  (as  in  dams)  and  thus 
transforming  it  into  potential  energy,  and  by  controlling  the  expenditure 
of  kinetic  energy  with  small  variations  of  potential  energy  (as  on  the  grid 
of  the  electronic  tube) . 

Neutralization  is  considered  to  be  the  result  of  structure  building 
by  the  process  of  binding  [see  274,  and  268,  particularly  part  7].  These 
structures,  by  raising  the  discharge  thresholds  of  drive  energies  and  by 
building  new  controlling  "dams,"  obstruct  the  tendency  toward  direct 
discharge,  enforce  delay  and  detour,  and  thus  give  rise  to  derivative  moti- 
vations whose  tendency  toward  immediate  and  direct  discharge  is  de- 
creased :  thus,  a  step  toward  the  neutralization  of  cathexes  is  made.  With 
further  structure  building,  further  derivative  motives  arise,  which  expend 
cathexes  of  an  even  higher  degree  of  neutralization. 

Observations  also  necessitate  the  assumption  of  transformations  which 
reverse  the  effects  of  binding  and  neutralization.  These  observations  per- 
tain on  the  one  hand  to  the  weakening  of  controls  and  defenses,  and  on 
the  other  to  the  so-called  libidinization  or  aggressivization  of  functions 
and  structures.  These  transformations  may  be  termed  "mobilization35  or 
"deneutralization."83  Referents  of  these  transformations  may  be  observed 
in  special  normal  states  (e.g.,  dreams),  in  pathological  conditions  (e.g., 
compulsions  and  delusions),  etc. 

The  complexity  of  this  theory  of  psychological  energies  and  of  their 
relationship  to  the  motor  energy  of  behavior  has  far-reaching  conse- 
quences for  quantification.  The  motor  energy  of  behavior  is  "controlled" 
and  "released"  by  the  economics  of  psychological  energies  and  by  the 
corresponding  dynamics  of  the  psychological  forces  which  operate  through 
psychological  structures.  It  might  be  suggested  that  this  relationship  is 
akin  to  the  control  of  large  amounts  of  energy  (muscular)  by  an  in- 
formation network  operating  with  smaller  amounts  of  energy  (psy- 

83  They  occur  in  the  process  of  regression.  Freud  discusses  them  as  the  dissolu- 
tion by  regression  of  the  fusion  of  libidinal  and  aggressive  drives  (defusion)  [see 
131,  pp.  46-48;  see  also  148,  chapter  on  "The  Metapsychology  of  Hypnosis  and 


chological).  In  turn,  within  the  range  of  psychological  energies  we  find  a 
similar  relationship :  the  expenditure  of  large  amounts  of  energy  ( mobile 
drive  energy)  is  controlled  by  a  network  operating  with  smaller  amounts 
of  energy  (bound  and  neutralized  energy) .  Moreover,  since  the  processes 
of  binding  (structure  building)  and  neutralization  recur,  creating  ever 
new  layers  of  the  ascending  hierarchy  of  psychological  organization,  we 
are  faced  with  a  whole  array  of  controlling  networks  arranged  in  depth. 

One  of  the  obstacles  to  quantification  now  becomes  obvious.  Overt 
behavior  is  as  a  rule  a  remote  representation  of  the  psychological 
processes  which  give  rise  to  it.  Thus,  even  though  observations  strongly 
suggest  the  need  for  concepts  like  cathexis  and  those  referring  to  cathectic 
transformations,  the  measurement  of  their  referents  is  made  difficult  by 
the  circumstance  that  the  same  set  of  observations  necessitates  the  as- 
sumption of  a  complex  hierarchical  control  organization  regulating 
cathectic  expenditure  and  transformation  by  means  of  structures  operat- 
ing with  small  amounts  of  cathexis.  If  the  theory  were  systematically 
tight,  its  definitions  explicit,  and  its  implicative  rules  specific,  the  dis- 
continuities, resulting  from  the  multiple  controls  which  cathectic  ex- 
penditures are  subject  to,  would  not  obstruct  quantification.  But  the 
theory  is  far  from  being  that  tightly  knit.  The  best  index  of  the  theory's 
looseness  is  that  the  volume  of  its  experimentally  verified  propositions 
would  be  ample  to  confirm  a  tighter  theory. 

Yet  the  situation  is  not  as  hopeless  as  the  complexities  described 
suggest.  The  theory  of  cathexes  does  include  quasi-quantitative  proposi- 
tions in  the  form  of  inequalities.  For  instance,  the  following  inequality 
holds  for  mobile  cathexes:  in  drive  action,  the  quantity  of  cathexis  is 
greater  than  that  in  affect  charge,  which  in  turn  is  greater  than  that  in 
an  idea.  Such  a  series  of  inequalities  is  per  se  a  primitive  (intensive) 
quantification84  and  this  kind  of  quantification  is  inherent  in  the  theory7. 
For  instance,  there  is  no  reason  why  the  degrees  of  mobility  of  cathexis 
(or  conversely,  the  degrees  of  neutralization)  should  not  be  expressed  in 
terms  of  such  inequalities.  Indeed,  R.  R.  Holt's  [178]  study  of  primary 
process  manifestations  in  the  Rorschach  test  did  just  that.  Such  ordinal 
scaling,  using  the  psychologist's  rating  procedures,  seems  for  the  present 
the  quantification  method  of  choice  for  the  primitive  quantitative  rela- 
tionships of  the  theory.  Some  of  its  difficulties,  however,  should  be  men- 
tioned here : 

Ordinal  scaling  of  primary  process  phenomena  may  distract  attention 
from  the  fact  that  the  theory  does  not  posit  a  simple  continuum  of 
neutralization.  The  decrease  of  mobility  goes  along  with  binding  (struc- 

84  Cf.  Piaget's  [252]  discussion  of  intensive  quantification.  Altogether,  Piaget's 
discussions  of  the  development  of  quantity  concepts  and  its  relation  to  logic  and 
mathematics  are  relevant  to  the  quantification  problems  of  psychology. 

The  Structure  of  Psychoanalytic  Theory  129 

ture-building),  that  is,  with  the  establishment  of  new  hierarchic  levels 
which  differ  from  each  other  not  only  in  the  degree  of  mobility  of 
cathexis,  but  also  in  their  structures  and  in  their  kinds  of  motivations. 
While  the  degree  of  mobility  remains  a  common  parameter  through- 
out the  various  hierarchic  levels,  the  qualitative  differences  in  structures 
and  motivations  from  level  to  level  make  it  difficult  to  find  that  feature 
of  behavior  which,  when  rated  and  scaled,  will  quantify  that  common 
parameter.  Whenever  the  behavior  feature  chosen  for  rating  is  not 
appropriate,  apples  and  pears  will  be  compared.  R.  R.  Holt  seems  to 
have  avoided  this  pitfall,  but  it  required  a  considerable  mastery  of  the 
theory  to  do  so,  and  the  measures  used  remained  gross. 

Thus,  individual  instances  of  primary  process  phenomena  do  not 
offer  an  insurmountable  obstacle  to  quantification.  But  how  about  a  be- 
havior segment  (a  Rorschach  or  a  TAT  record,  or  a  clinical  interview) 
which  contains  several  such  instances?  Each  of  these  can  be  rated.  But 
may  we  count  them?  If  we  do,  what  is  our  justification  for  doing  so? 
If  we  decide  to  weight  them,  are  the  weights  additive?  These  questions  are 
not  yet  answered.  We  do  not  even  know  where  the  answers  will  come 
from.  We  may  have  to  accept  purely  empirical,  theoretically  unsupported 
answers  for  a  long  while,  in  the  hope  that  these  will  show  us  that  the 
theory7  has  (or  can  be  expanded  to  have)  the  answers.  It  is  also  possible 
that  the  empirical  answers  will  radically  change  the  theory.  This  problem 
is  not  specific  to  Holt's  study.  In  food  deprivation  studies  using  TAT 
stories,  we  find  individual  differences  in  the  stories  of  a  group  of  equally 
deprived  subjects:  some  stories  contain  much  material  distantly  related 
to  food,  others  contain  little  food-related  material,  but  what  there  is,  is 
closely  related.  Can  the  ratings  of  these  individually  differing  products 
be  added  up?  The  relationship  between  drive  intensity  (amount  of 
cathexis)  on  the  one  hand  and  the  frequency  and  intensity  of  its  in- 
dicators on  the  other  is  a  significant  unsolved  problem  of  quantification. 

B.  Dimensional  Quantification85 

What  are  the  general  prospects  for  the  quantification  of  the  variables 
of  this  theory? 

Before  attempting  an  answer  to  this  question,  let  us  state  that  the 
urgent  tasks  of  this  theory  are  in  the  relationships  it  posits  which  re- 
quire systematization,  and  in  the  areas  which  require  new  observations. 
Without  stressing  that,  among  other  things,  much  of  ego  psychology  is 
still  uncharted  territory,  and  that  our  knowledge  of  affects  is  in  urgent 
need  of  systematization,  etc.,  the  very  discussion  of  quantification  may 

85  Since  the  completion  of  this  manuscript,  A.  Menkes  and  J.  Menkes  have 
published  a  paper  [231]  which  contains  an  example  of  this  kind  of  quantification 
and  goes  a  considerable  way  in  demonstrating  the  necessity  for  such. 


misrepresent   the   actual   situation:    mathematization   in   general    and 
quantification  in  particular  require  a  systematized  and  tightly  knit  theory. 

When  the  physicist  measures,  he  knows  the  dimensions  of  his  ob- 
servables  as  expressed  in  terms  of  the  CGS  (centimeter,  gram,  second) 
system,  and  when  he  establishes  a  constant  he  knows  that  its  dimension 
is  such  as  to  make  his  equation  not  only  quantitatively  but  also  di- 
mensionally  true.  In  s  =  (g/2)t2  the  dimension  of  s  is  C,  of  t  is  S,  and 
of  g  is  C/S2;  thus,  substituting  these  dimensions,  we  get  C  =  (C/S2}S2, 
indicating  that  the  equation  is  dirnensionally  true.  The  classic  scale  of 
hardness  is  a  means  of  quantification  too.  But  instead  of  a  dimensional 
measure,  it  provides  only  an  ad  hoc  quantification.  Most — if  not  all — 
measurements  (e.g.,  IQ's)  of  present-day  psychology  are  ad  hoc  quanti- 
fications. Without  a  systematized  theory,  no  dimensional  quantification  is 
possible.  In  physics,  nobody  would  try  to  test  a  theory  by  a  measurement 
without  first  ascertaining  the  relevance  of  what  he  measures  and  how  he 
measures  it.  The  dimensions  are  the  criteria  of  relevance.  Psychologists, 
however,  "test"  psychoanalytic  propositions  without  studying  and  system- 
atizing the  theory  which  gives  meaning  to  these  propositions.  Theories 
can  be  tested  only  when  they  are  taken  seriously.  To  test  is  to  mathema- 
tize  and  to  mathematize  is  to  discover,  in  the  relationships  posited  by  the 
theory,  relationships  of  a  higher  order  of  abstraction.  Such  abstractions 
cannot  be  derived  from  isolated  propositions,  but  only  from  the  system 
of  relationships  which  link  these  to  each  other. 

So  far  we  do  not  know  how  to  achieve  a  dimensional  quantification 
of  psychoanalytic  variables;  and  yet  we  cannot  sit  with  folded  hands, 
since  additional  observations  are  needed  for  the  systematization  of  the 
theory  and  for  dimensional  quantification.  Thus  in  gathering  new  ob- 
servations we  must  be  satisfied  with  ad  hoc  quantifications,  but  we  must 
not  lose  sight  of  the  goal  of  dimensional  quantification.  To  achieve  that, 
we  will  have  to  learn  to  consider  the  locus  of  our  variables  in  the  motiva- 
tional and  structural  hierarchy  and  to  play  variables  against  each  other 
so  as  to  arrive  at  equations  which  represent  actual  balances  of  forces,  or 
balances  between  structures  and  forces,  etc.  Progress  toward  dimensional 
quantification  will  at  every  step  require  long  series  of  experiments  which 
vary  the  experimental  conditions  systematically.  The  currently  fashion- 
able one-shot  experiments  (probably  fostered  both  by  the  premium  put 
on  publication  and  by  the  publication  policy  of  psychological  journals) 
militate  against  progress  toward  dimensional  quantification.  One-shot 
experiments,  naturally  enough,  use  ad  hoc  quantifications,  and  only 
rarely  cogwheel  into  the  ad  hoc  quantifications  of  other  experiments. 
Lewinian  experiments  in  affect  and  action  psychology  avoided  this  pitfall 
to  some  extent  and  showed  how  ad  hoc  relationships  can  be  avoided  by 
systematic  variation  of  experimental  conditions  directed  by  a  cohesive 

The  Structure  of  Psychoanalytic  Theory  131 

theory.  But  the  reports  of  these  experiments  are  In  German  and  thus 
have  been  little  read,  except  in  Ellis's  "52]  excerpts  or  Lewin's  [216] 
summary,  neither  of  which  conveys  the  method. 

One  of  the  banes  of  ad  hoc  quantification  is  that  even  when  it  yields 
statistically  reliable  results,  these  may  be  due  to  sheer  luck  in  choosing 
the  experimental  tasks  and  subjects.  Even  the  apparently  precise  replica- 
tion of  an  experiment  may  bring  different  results.  The  crucial  dimensions 
not  being  known,  unnoticed  "minor33  variations  of  the  setup  affect  the 
results.  In  other  words,  without  knowing  the  dimensions  involved  it  is 
impossible  to  predict  what  changes  will  make  for  hierarchic  differences 
and  for  what  types  of  subjects  will  the  objectively  "precise"  replication 
amount  to  a  radically  different  setup.  George  Klein  [197,  200]  has  shown 
that  something  of  this  sort  was  involved  in  the  "now  you  see  it,  now  you 
don't53  character  of  the  Bruner-Goodman  [37]  effect. 

Now,  as  to  the  possibility  and  prerequisites  of  dimensional  quanti- 
fication: First,  dimensional  quantification  in  psychology  may  not  be 
feasible.  We  would  be  reluctant  to  entertain  this  possibility,  partly  be- 
cause it  would  require  negative  proof,  which  is  difficult  if  not  impossible 
to  obtain,  and  partly  because  it  would  discourage  further  research. 
Second,  the  quest  for  dimensional  quantification  may  lead  to  a  nonrnetric 
mathematization.  Third,  a  dimensional  quantification  may  develop. 

The  avenue  by  which  we  may  be  able  to  arrive  at  a  choice  among 
these  three  possibilities  will — to  my  mind — be  paved  by  a  new  start  on 
the  problem  of  learning: 

The  physical  dimensions  are  mass,  space,  and  time.  Physics  ex- 
presses both  the  movement  of  mass  (i.e.,  its  changes  of  position  in  space 
and  time)  and  the  changes  in  the  structure  of  mass,  as  well  as  the 
gradients  and  causes  (force,  energy)  of  such  changes,  in  terms  of  these 
dimensions.  If  we  were  to  have  psychological  dimensions,  they  too 
would  have  to  be  able  to  express  psychological  processes  as  well  as 
psychological  structures  and  their  changes.  In  psychoanalytic  theory, 
structures  play  such  a  crucial  role  that  as  long  as  the  propensities 
and  changes  of  psychological  structure  cannot  be  expressed  in  the  same 
dimensions  as  psychological  processes,  dimensional  quantification  is  but  a 
pious  hope.  In  other  words,  the  study  of  the  process  of  psychological 
structure  formation  seems  to  be  the  prime  requisite  for  progress  toward 
dimensional  quantification.  We  must  establish  how  processes  turn  into 
structures,  how  a  structure,  once  formed,  changes,  and  how  it  gives  rise 
to  and  influences  processes.  This  could  be  achieved,  for  instance,  by 
studying  the  processes  by  which  Hebb's  hypothetical  structures  (as- 
semblies and  phase  sequences)  are  formed  and  changed,  as  well  as  the 
processes  by  which  these  structures  change  new  ongoing  processes.  Like- 
wise, this  could  be  achieved  by  the  study  of  those  structures  whose  genesis 


and  function  psychoanalysis  conceives  of  as  follows:  when  drives  en- 
counter an  obstacle  to  the  discharge  of  their  cathexes,  structures  are 
formed  and  these  structures  thereafter  serve  both  as  obstacles  to  (de- 
fenses against)  and  controls  and  means  of  discharge.  These  examples 
refer  to  changes  wrought  by  experience.  Whether  or  not  all  structure 
formation  (in  that  broad  sense  which  takes  account  of  the  epigenetic- 
maturational  matrix)86  should  be  considered  learning  (i.e.,  abiding 
change  wrought  by  experience)  is  both  an  empirical  and  a  conceptual 
problem.  But  it  seems  that  all  learning  may  be  looked  upon  as  a  process 
of  structure  formation.  The  processes  of  verbal  learning  and  habit 
formation  may  well  be  considered  subordinate  to  this  broader  category, 
though  their  study  may  or  may  not  be  revealing  of  the  relationship  be- 
tween process  and  structure. 

What  study  will  reveal  this  relationship?  Thirty  years  ago,  Adams  [2] 
suggested  that  the  main  obstacle  in  the  way  of  the  study  of  the  learning 
process  is  its  slowness.  Hebb  [169]  pointed  to  the  slow  rate  of  early 
learning  processes.  The  burden  of  Piaget's  [254,  255,  256,  etc.]  develop- 
mental studies  in  and  since  The  Origins  of  Intelligence  is  the  same. 
It  is  possible  that  only  longitudinal  studies  can  clarify  the  relationship 
between  process  and  structure.  But  since  the  methodology  of  longitudinal 
studies  is  still  obscure,  this  is  a  dim  hope.  Perhaps  the  answer  will  come 
from  a  new  attack  on  learning  as  structure  formation,  which  will  take 
account  of  Hebb's  assumption  that  late  learning  operates  by  recombining 
already  established  "phase  sequences' 3  and  will  thus  center  on  the  changes 
in,  rather  than  on  the  origin  of,  such  phase  sequences. 

The  immediate  outlook  for  an  early  clarification  of  the  process  of 
structure  formation  seems  none  too  rosy.  Yet  this  clarification  appears 
to  be  the  prerequisite  for  dimensional  quantification  in  psychoanalysis 
in  particular,  and  perhaps  even  in  psychology  at  large. 

But  the  quest  for  dimensional  quantification  must  not  amount  to 
a  disdain  for  ad  hoc  quantification.  The  latter  seems  to  be  a  step 
toward  the  former,  provided  it  is  clearly  understood  that  ad  hoc  quanti- 
fication itself  does  not  locate  hierarchically  the  structures  and  functions 
which  it  crudely  quantifies.  The  possibility  of  arriving  at  a  dimensional 
quantification  can  be  kept  open  by  matching  the  care  and  ingenuity  ex- 
pended on  ad  hoc  quantifications  with  an  unremitting  alertness  for  the 
hierarchic  locus  of  the  relationships  so  quantified. 

All  this  discussion  of  quantification  is,  however,  in  a  sense  abstract 
and  sterile.  A  proper  discussion  would  have  to  start  out  with  an  analysis 
of  the  experimental  literature  pertaining  to  Freudian  propositions.  We 
have  several  surveys  of  this  literature  [e.g.,  308,  173],  but  their  concern 
is:  what  psychoanalytic  propositions  are  confirmed  by  "objective  studies?" 
86  Cf.  pp.  86-88,  above. 

The  Structure  of  Psychoanalytic  Theory  133 

A  survey  which  could  advance  the  solution  of  the  mathematization  or 
quantification  problem  would  have  to  center  not  on  the  results  of  these 
studies  but  on  their  method,  on  the  variables  which  were  the  targets 
of  quantification  in  them,  and  on  the  technique  of  quantification  used  by 
them.  Short  of  a  breakthrough  by  means  of  experimental  ingenuity 
coupled  with  thorough  theoretical  grounding,  such  a  survey  seems  to 
hold  the  best  promise  of  progress  toward  the  solution  of  the  quantification 


A.  The  Present  Status  of  the  System 

The  theory  of  psychoanalysis  grew  by  successive  spurts  in  the  fifty 
years  of  Freud's  work.  Additions  and  revisions  make  it  appear  more 
like  a  patchwork  than  an  architectonic  design,  since  their  consequences 
for  the  structure  of  the  system  have  often  remained  a  matter  of  a  passing 
comment  by  Freud  or  isolated  papers  by  other  psychoanalysts.  This  in 
itself  bespeaks  a  looseness  of  the  theory  and  its  lack  of  an  explicit  canon 
according  to  which  revisions  and  additions  are  to  be  fitted  into  its  system. 
Yet  psychoanalytic  theory  does  have  an  impressive  structural  unity, 
though  it  is  hidden  under  the  layers  of  progressive  additions  and  modi- 
fications, and  has  not  been  disentangled  and  independently  stated. 

The  "revisors"  of  Freud's  theory  further  obscured  its  structural  unity. 
Jung  and  Adler,  who  created  relatively  independent  theories,  failed  to 
give  these  a  systematic  form  which  could  have  sharply  distinguished  them 
from  Freudian  theory.  The  situation  is  even  worse  with  the  "revisions33  of 
Stekel,  Rank,  Sullivan,  Homey,  M.  Klein,  Kardiner,  Alexander,  French, 
Reik,  Fromm,  Rado  [see  Munroe,  240] .  While  each  attacked  and  denied 
certain  Freudian  propositions,  and  replaced  them  by  others  (which 
often  contained  a  valid  core),  none  of  these  authors  stated  how  their  re- 
visions affect  the  theory  as  a  whole.  Some  of  them  (Stekel,  Kardiner, 
Alexander,  French,  Reik)  have  asserted  that  their  revisions  do  not  affect 
the  rest  of  the  system,  although  they  made  no  attempt  to  demonstrate 
this.  Others  (Rank,  Horney,  Sullivan,  Rado)  have  implied  that  Freud's 
system  has  been  replaced  by  their  own,  although  they  never  presented  a 
full  elaboration  of  their  systems.  No  Neo-Freudian  has  taken  cognizance 
of,  and  has  integrated  his  own  contribution  with,  the  whole  of  psycho- 
analytic theory.  Nor  is  there  a  single  attempt  to  replace  it  with  a  whole 
system  that  demonstrably  accounts  for  all  the  phenomena  psychoanalytic 
theory  claims  to  explain.  Such  an  attempt  could  obviously  include  a 
demonstration  that  some  of  the  problems  psychoanalytic  theory  dealt 
with  are  pseudoproblems  which  can  be  ignored.  The  lack  of  an  explicit 


statement  of  the  theory  is  as  much  responsible  for  aH  this  as  are  the 
"revisors"  themselves,  who  may  have  felt  that  they  were  not  obliged  to 
disentangle  the  system  before  they  re\ised  it.  Study  of  Neo-Freudian 
writing  often  makes  one  wonder  whether  the  authors  were  aware  of  the 
existence  and  nature  of  the  implicit  system  of  psychoanalytic  theory. 

There  are  three  outstanding  rudimentary1  statements  of  the  theory's 

First 9  Freud's  seventh  chapter  of  The  Interpretation  of  Dreams  [98] 
and  his  "Papers  on  Metapsychology"  [108,  110,  114,  115,  116,  117,  119, 
120]  are  attempts  to  present  the  system.  One  of  the  most  puzzling  prob- 
lems of  the  history  of  psychoanalysis  is  why  they  were  so  little  noticed. 
The  fact  that  the  form  of  these  attempts  is  not  systematic  does  not  seem 
to  explain  this  fully.  The  formulations  of  the  present  essay  derive  from 
these  writings,  and  so  do  the  other  attempts  at  systematization  to  be 
mentioned  here. 

Second,  FenichePs  The  Psychoanalytic  Theory  of  Neurosis  [73], 
while  it  is  focused  on  the  clinical  theory  of  psychoanalysis,  does  per- 
sistently invoke  the  general  theory  and  thus  gives  a  sense  of  its  system. 
Yet  the  latter  remains  implicit,  and  the  experimenter  who  wishes  to  start 
from  FenichePs  formulations  must  first  disentangle  them  from  their 
clinical  matrix.  With  FenichePs  death,  psychoanalysis  lost  one  of  its  few 
systematizers.  His  essay  on  the  theory  of  technique  [72]  is  a  beginning  of 
the  systematization  of  the  theory  of  therapy.  His  posthumously  published 
Collected  Papers  [77]  contains  systematic  discussions  of  M.  Klein,  Kaiser, 
Fromm,  and  other  "revisionists."  These  discussions,  as  well  as  his  paper 
on  Freud's  theory  of  the  death  instinct,  show  that  the  psychoanalytic 
theory  is  sufficiently  cohesive  to  permit  systematic  exclusion  and  in- 
clusion of  new  contributions. 

Third,  the  development  of  ego  psychology  is  perhaps  the  clearest 
demonstration  of  the  systematic  nature  of  psychoanalytic  theory.  In 
Anna  Freud's  [93]  work  the  clinical  theory  of  defenses  begins  to  take 
a  systematic  form;  in  Erikson's  [61]  work  the  development  of  the  ego 
and  the  psychosocial  theory  of  psychoanalysis  takes  shape;  and  in  Hart- 
mann's  [157]  work  [complemented  by  Kris's  and  Loewenstein's,  167, 
168,  206]  the  theory  of  the  ego  develops  hand  in  hand  with  a  progressive 
crystallization  of  the  general  theory  of  psychoanalysis.  All  these  con- 
tributions show  that  psychoanalytic  theory  can  grow  organically  so  as  to 
include  the  valid  observations  and  formulations  of  the  Neo-Freudians, 
without  becoming  an  incoherent  patchwork  and  without  the  necessity 
of  discarding  any  of  its  major  segments.  They  demonstrate  that  the 
theory  has  sufficient  systematic  coherence  not  only  to  reject  incompatible 
solutions,  but  also  to  develop  compatible  theories  of  the  ego,  of  reality, 
of  interpersonal  relationships,  and  of  social  psychology. 

The  Structure  of  Psychoanalytic  Theory  135 

B.  The  Desirable  Level  of  Formalization 

The  desirable  level  of  fomialization  is,  in  a  sense,  an  empirical  ques- 
tion. Since  everybody  wants  to  be  on  the  side  of  the  angels,  we  may  as- 
sume that  reaching  maximal  explicitness  is  the  ideal  and  only  the  limita- 
tions of  our  knowledge  stand  in  the  way.  Newton's  axiomatization  was 
explicit  and  its  heuristic  value  shows  that  it  was  desirable.  But  the 
systematic  and  heuristic  value  of  Einstein's  last  formalizations  is  much 
questioned.  Present-day  physics  has  no  unified  axiomatic  system.  All  in 
all,  probably  only  experience  can  decide  when  and  how  far  axiomatiza- 
tion can  be  meaningfully  pushed  in  psychology  or  in  any  other  science. 

Yet  it  may  be  worth  while  to  raise  the  question:  why  are  psy- 
chologists so  concerned  with  axiomatization?  Actually,  axiomatization 
has  always  been  a  late  product  in  every  science.  Centuries  of  Egyptian 
geometry  preceded  Euclid.  Newton  had  not  only  Galileo  and  Kepler,  but 
thousands  of  years  of  physics  behind  him.  Sciences  do  not  arise  from,  but 
culminate  in,  axiomatics.  Axiomatic  systems  do  not  reveal  the  tracks  of  a 
science's  development:  they  conceal  them.  They  do  what  so  many  psy- 
chologists do  who  arrive  at  their  results  with  great  difficulty  (like  the  rest 
of  us),  but  from  reading  their  published  papers  one  would  never  guess 
that;  they  seem  to  reveal  a  foresight  which  puts  to  shame  all  others  who 
deal  with  human  beings  or  govern  human  affairs. 

Does  the  yearning  for  axiomatization  mean  that  psychologists  believe 
psychology  can  arrive  at  its  future  by  lifting  itself  by  its  own  bootstraps? 
Are  we  really  to  believe  that  we  can  guess  our  way  through  to  axiomatics 
and  bypass  the  long  road  other  sciences  have  had  to  travel  toward  it? 
Beat  the  other  sciences  at  the  game?  Or  just  simply  profit  by  their  ex- 
perience? But  what  if  our  guesses  lead  to  a  disregard  of  the  empirical 
evidence  we  already  have  and  to  a  lack  of  concern  for  the  evidence  that 
is  not  yet  in?  What  if  the  attempts  at  short-cutting  the  arduous  path  of 
development  lead  only  to  endless  detours — much  longer  than  the  "long 
and  hard"  empirical  route,  and  futile,  to  boot?  Is  it  possible  that  psy- 
chologists ignore  what  the  natural  scientist  [24]  and  the  historian  of 
science  [46]  have  come  to  recognize :  that  scientific  discovery  starts  from 
intuition  and  not  from  deduction? 

This  is  not  to  question  that  psychology  can  profit  by  the  experience 
of  other  sciences,  nor  to  make  light  of  axiomatics  as  an  ideal,  nor  to 
minimize  its  importance  in  the  development  of  sciences,  nor  to  contend 
that  theory  making  (including  axiomatics)  is  not  as  essential  to  science 
as  "measurement33:  intuition  or  hunch  is  theory.  The  point  is  that  in 
present-day  psychology  the  measuring  furor  seems  to  have  made  an 
unholy  alliance  with  an  axiomatic  -furor,  and  between  the  two  of  them 
they  may  well  doom  psychology  to  stagnation. 


Piaget  in  his  Epistemologie  Genetique  [252]  examined  both  the 
history  of  mathematics  and  the  ontogenetic  development  of  mathematical 
reasoning  in  an  attempt  to  explain  how  mathematics  can  be  simul- 
taneously deductively  rigorous  and  yet  fertile.  The  study  of  his  investiga- 
tions is  a  good  antidote  to  premature  axiomatization. 

Psychoanalysis  is  in  sore  need  of  systematization,  because  without 
it  the  experimenter  is  likely  to  continue  to  test  isolated  and  misconstrued 
propositions,  unaware  of  their  actual  theoretical  context.  But  systematiza- 
tion is  a  long  way  from  formalization  and  axiomatization.  Much  addi- 
tional knowledge  will  have  to  accumulate  before  we  can  even  begin  to 
work  on  the  latter  tasks. 


The  theory,  though  it  originated  in  the  study  of  pathology,  has 
always  claimed  to  explain  normal  behavior  and  development  also — 
Psychopathology  of  Everyday  Life  [99],  Wit  and  Its  Relation  to  the  Un- 
conscious [100],  "Humour"  [134],  Three  Essays  on  the  Theory  of 
Sexuality  [101]. 

Moreover,  Freud  demonstrated  that  the  theory  and  its  methods 
can  be  fruitfully  applied  to  anthropology  and  prehistory  [Totem  and 
Taboo,  109];  to  the  study  of  literature  [Delusion  and  Dream,  102,  "The 
Relation  of  the  Poet  to  Day-dreaming,33  103,  "Dostoevsky  and  Parri- 
cide,53 133,  "The  Theme  of  the  Three  Caskets,"  1 12] ;  to  the  study  of  art 
[Leonardo  da  Vinci,  105,  "The  Moses  of  Michelangelo,"  113];  to  the 
study  of  mythology,  folklore,  and  legend  [Totem  and  Taboo,  109,  "A 
Mythological  Parallel  to  a  Visual  Obsession,"  118,  "Medusa's  Head," 
125,  "The  Occurrence  in  Dreams  of  Material  from  Fairy-tales,"  111]; 
to  the  study  of  language  ["The  Antithetical  Sense  of  Primal  Words," 
104];  to  the  study  of  religion  [Totem  and  Taboo,  109,  "A  Religious  Ex- 
perience," 135,  The  Future  of  an  Illusion,  132,  Moses  and  Monotheism, 
141];  to  the  study  of  history  [Moses  and  Monotheism,  141];  and  to  the 
study  of  society  [Totem  and  Taboo,  109,  Group  Psychology  and  the 
Analysis  of  the  Ego,  124,  Civilization  and  Its  Discontents,  136,  "Why 
War?"  138].  Finally,  Freud  at  various  times  asserted  the  applicability 
of  his  method  and  theory  to  those  phenomena  which  we  subsume  under 
the  term  psychosomatics. 

Indeed,  Freud  considered  all  human  behavior  and  endeavor  to  be 
within  the  purview  of  psychoanalysis.  Psychoanalysts  followed  his  lead 
and  their  literature  abounds  in  papers  and  books  dealing  with  the  fields 
listed.  Even  though  these  contributions  aroused  heated  and  often 
acrimonious  debate,  and  even  though  their  cogency  and  their  scholarship 
in  the  field  in  question  have  been  criticized  sharply  and  often  rightly, 

The  Structure  of  Psychoanalytic  Theory  137 

the  present  situation  in  all  of  these  fields  seems  to  bear  out  Freud's  early 

In  summary,  psychoanalytic  theory  has  asserted  an  all-inclusive  ap- 
plicability to  the  study  of  man.  Psychoanalysts  have  acted  to  make  this 
claim  good.  Investigators  in  the  various  fields  approached  by  psycho- 
analysis have  adopted  some  of  its  methods,  concepts,  theories,  and  out- 
look. There  is  sufficient  evidence  that  the  claim  has  a  substantially  valid 

Now  we  come  to  the  applications  of  psychoanalysis  to  psychology 
proper.  Here  we  can  give  only  a  brief  sketch  of  the  complex  problems 
involved  [see  Shakow  and  Rapaport,  309]. 

Though  Freud  conceived  of  psychoanalysis  as  a  general  psychology, 
little  in  his  theory  pertained  directly  to  psychophysics,  learning,  and  per- 
ception, the  areas  central  to  academic  psychology,  and  he  did  not  at- 
tempt to  apply  his  theories  or  methods  to  psychology  at  large. 

At  first  only  a  few  psychoanalysts  showed  an  interest  in  psychology: 
for  example,  Schilder  [299,  300,  301,  302],  Bernfeld  [13],  de  Saussure 
[297].  But  through  developmental  psychology  [Piaget,  247,  248,  249, 
250,  251,  and  Werner,  322],  through  early  clinical-experimental  psy- 
chology [Murray,  242,  243,  and  Rosenzweig,  293,  294],  through  the 
influence  of  projective  techniques  on  clinical  psychology  [Rorschach,  292, 
Morgan  and  Murray,  238,  and  others],  through  learning  theory  [Dollard 
and  Miller,  48;  HuU,  183;  Mowrer,  239], 8T  and  through  psychologists3 
growing  interest  in  psychotherapy,  psychoanalysis  came  to  exert  a  power- 
ful influence  on  psychology  proper.  Most  of  this  influence  did  not  stem 
from  psychoanalysts'  applying  their  theory  and  methods  to  psychology 
(Jung,  Rorschach,  and  Murray  may  be  considered  exceptions)  but 
rather  from  psychologists'  attempting  to  use  the  conceptions  (rather  than 
the  concepts)88  of  psychoanalysis. 

Only  with  the  development  of  psychoanalytic  ego  psychology  did 
psychoanalysis  begin  to  acquire  means  for  dealing  with  the  usual  prob- 
lems of  psychology,  Hartmann  [157]  then  made  it  explicit  that  psycho- 
analysis is  a  general  psychology,  that  its  interest  and  application  extend 
to  the  field  of  academic  psychology,  and  proceeded  to  link  psychoanalytic 
and  psychological  propositions  to  each  other.  Subsequently,  several  psy- 
choanalysts and  psychoanalytically  trained  psychologists  continued  to 
relate  psychological  and  psychoanalytic  conceptions,  theories,  concepts, 

8TFor  earlier  psychoanalytic  influence  on  learning  theory,  see  E.  B.  Holt  [174, 
175],  Kempf  [196],  Humphrey  [184,  185],  Troland  [318],  and  others. 

88  Concepts  are  terms  defined  within  the  framework  of  a  theory,  conceptions 
are  terms  and  formulations  which  either  precede  the  definition  of  the  concepts  in 
the  history  of  a  theory  or  disregard  them.  Thus  statements  of  conceptions  use  the 
terms  of  a  theory  in  an  imprecise  or  arbitrary  "common-sense"  fashion. 


and  methods  to  each  other.  It  is  to  be  hoped  that,  as  a  consequence  of 
this  work,  the  haphazard  "experimental  testing"  of  psychoanalytic 
theories  and  their  untested  application  by  clinical  psychologists  may 
eventually  give  way  to  their  systematic  application  to  psychology7,  within 
the  framework  which  psychoanalytic  ego  psychology  has  begun  to  build. 
To  advance  this  work  of  mutual  application,  the  theory  of  psychoanalysis 
must  face  two  major  tasks  besides  systematization :  coming  to  terms  with 
Piagefs  theory  and  developing  a  learning  theory. 

If  Piagefs  [252,  254,  255,  256]  findings  are  confirmed,  psycho- 
analysis will  have  to  come  to  terms  with  his  developmental  theory  as  an 
indispensable  segment  of  the  theory  of  ego  development.  The  problems 
to  be  solved  before  this  can  be  accomplished  cannot  be  sketched  here.89 

Our  discussion  of  learning  (see  Section  V.  B.)  suggested  that 
dimensional  quantification  may  not  be  possible  without  a  prior  clarifica- 
tion of  the  process  of  structure  formation  and  learning.  But  the  solution 
of  the  problem  of  structure  formation  may  also  be  one  of  the  pre- 
requisites for  a  unified  theory  of  cognition  (including  perception),  for  the 
clarification  of  the  methodology  of  developmental  studies,  and  perhaps 
for  the  solution  of  still  other  issues  crucial  both  for  the  systematic  de- 
velopment of  psychoanalysis  and  for  the  mutual  and  fertile  application 
of  psychological  and  psychoanalytic  methods  and  theories. 


It  would  take  volumes  to  sketch  and  critically  appraise  all  the 
research  that  has  been  "mediated"  by  psychoanalytic  theory,  by  hunches 
derived  from  it,  questions  raised  by  it,  and  methods  originating  in  it. 
It  is  not  feasible  to  list  even  the  highlights  of  such  research  in  the 
fields  mentioned  in  the  previous  section.  Therefore,  we  will  restrict  our- 
selves to  a  cursory  survey  of  its  research  mediation  in  psychodynamics 
and  psychology. 

There  are,  first  of  all,  the  clinical  studies  which  fill  the  psycho- 
analytic as  well  as  the  Neo-Freudian  literature.  Moreover,  it  has  been 
a  long  time  since  any  clinical  psychiatric  case  study  could  be  made  with- 
out drawing  on  psychoanalytic  theory,  which,  to  a  greater  or  lesser  ex- 
tent, has  mediated  much  of  what  passes  today  for  clinical  psychiatric  re- 
search. Finally,  the  psychosomatic  investigations  of  the  last  two  decades 
arose,  in  the  main,  from  psychoanalytic  studies  of  organ  neuroses,  were 
nursed  to  a  more  or  less  general  acceptance  by  the  work  of  psycho- 
analysts like  F.  Deutsch,  Alexander  and  French,  Dunbar,  and  Binger, 
and  have  been  turned  into  everyday  clinical  research  by  the  efforts  of 
Kubie,  Kaufman,  M.  Lewin,  Romano,  and  many  others. 

89  See   P.  H.  Wolffs  [328]   study  of  Piaget's  theory  and  his  discussion  of  its 
relation  to  psychoanalysis. 

The  Structure  of  Psychoanalytic  Theory  139 

A  selection  from  one  area  of  more  or  less  systematic  studies  which 
issued  from  psychoanalysis  proper  was  collected  and  reviewed  in 
Organization  and  Pathology  of  Thought  [268],  but  a  full  survey  of  all 
such  studies  has  not  yet  been  made. 

Projective  techniques,  which  have  come  to  play  an  increasing  role 
both  as  subject  matter  and  as  tools  of  psychological  research,  had  their 
origins  in  psychoanalytic  theory.  Rorschach  and  Murray  were  steeped  in 
psychoanalysis  and  their  tests  are  informed  by  psychoanalytic  conceptions. 
In  fact,  these  tests  came  into  clinical  use  carried  by,  and  carrying,  the  im- 
pact of  psychoanalytic  theory;  they  used  segments  of  that  theory  for 
their  rationale  and  interpretation  [176,  281];  and  they  wrere  used  to 
"test"  psychoanalytic  propositions.  Moreover,  both  these  and  play  tests 
(deriving  from  the  play  techniques  of  therapy)  bred  a  vast  array  of  newT 
projective  tests,  founded  on  and  "testing"  further  psychoanalytic  con- 
ceptions. How  valid  their  connection  to  and  their  "testing"  of  psycho- 
analytic assumptions  were,  need  not  concern  us  here:  "research"  was 

Throughout  the  last  forty  years,  psychoanalytic  theory  has  led  to 
an  extensive  array  of  experimental  studies  on  the  effect  of  emotions  and 
motivations  on  memory  [see  258].  Most  of  these  intended  to  test  the  psy- 
choanalytic theory  of  repression,  but  many  failed  to  distinguish  this  from 
hedonistic  pleasure-pain  theories  or  from  the  law  of  effect,  and  few  if 
any  were  really  conversant  with  it. 

A  related  area  of  research  mediated  by  psychoanalysis  is  that  of 
motivated  perception.  Murray  [242],  N.  Sanford  [295,  296]  pioneered, 
and  Murphy  and  his  pupils  [see  survey  in  241,  chap.  15]  continued  this 
line  of  investigation,  which  led  to  the  "new  look  in  perception,"  be- 
ginning with  Bruner's  [37,  38,  39,  40,  41]  work  and  reflected  in  Blake 
and  Ramsey's  [29]  volume.  Among  these,  from  the  point  of  view  of  this 
essay,  the  work  of  Klein  and  his  associatesS9a  stands  out.  While  all  these 
studies  bear  the  imprint  of  the  interest  in  motivation  aroused  in  psy- 
chology by  psychoanalysis  [cf.  Boring,  31,  pp.  693,  713],  the  Freudian 
influence  is  not  always  as  obvious  in  them  as  is  the  influence  of  Freud's 
motivation  theory  in  the  studies  of  Murray  and  Sanford,  and  that  of 
Freudian  motivation  and  ego  theory  in  the  work  of  Klein  and  his 

Psychoanalytic  theory  was  also  responsible  for  the  reawakening  of 
interest,  during  the  last  twenty  years,  in  the  nature  of  hypnosis  and  in  the 
use  of  hypnosis  as  an  experimental  method.  M.  Erickson's  [53,  54,  55], 
Farber  and  Fisher's  [68],  and  Gill  and  Brenman's  [33,  148]  hypnotic 
work,  as  well  as  Fisher's  [83]  work  with  waking  suggestion,  represent 
efforts  to  apply  psychoanalytic  theory  to  hypnosis  or  to  use  hypnosis  as  a 
means  of  psychoanalytic  exploration. 

89a  A  survey  of  these  will  be  found  in  Klein's  [198]  volume  soon  to  be  published 


Psychoanalysis  stimulated  and  guided  more  or  less  directly  many 
longitudinal  and  cross-sectional  developmental  studies  (Benjamin,  Esca- 
lona,  Spitz,  K.  Wolf,  and  others).  This  field  is  so  broad  that  neither  a 
further  listing  of  investigators  nor  a  bibliography  can  be  given  here.  But 
a  reference  to  Piaget  must  be  made.  Piagefs  early  work  (up  until  The  Ori- 
gins of  Intelligence,  254,  in  1935)  on  autistic  thinking  and  its  socializa- 
tion in  children  was  strongly  influenced  by  the  psychoanalytic  theory.90 
Piaget's  later  work  is  critical  of  psychoanalysis,  but  still  appears  to  show 
its  influence. 

Finally,  psychoanalysis — for  better  or  for  worse — has  also  mediated 
much  research  along  the  lines  of  the  learning  theories  which  originated 
at  Yale.  Whatever  view  one  takes  of  their  ultimate  pertinence  to  psycho- 
analytic theory,  Miller's  experimental  work  [234,  235,  236],  Miller  and 
Bollard's  studies  [237],  and  Mowrer's  experiments  [239],  as  well  as 
those  of  their  many  students,  certainly  arose  under  psychoanalytic  in- 

But  this  enumeration  of  major  areas  of  research  mediation  by  psycho- 
analysis in  psychology  disregards,  among  other  things,  social  psychologi- 
cal research  (e.g.,  on  authoritarianism)  and  does  not  do  justice  to  the 
pioneering  work  of  D.  Levy  [212,  213],  Halverson  [155,  156],  Murray 
[243],  J.  McV.  Hunt  [186,  187],  and  many  others.  Regrettably,  the 
existing  surveys— Sears  [307,  308],  Rapaport  [258],  Hilgard  [173]— 
are  either  specialized  or  incomplete.  A  careful  analytic  survey  of  the 
pertinent  experimental  literature  would  be  a  formidable  undertaking: 
the  amount  of  literature  on  research  purporting  to  be  related  and  on  re- 
search actually  related  to  psychoanalytic  theory  is  immense.  Yet  such  a 
survey  is  urgently  needed.  It  would  be  of  most  use  if  it  were  to  center 
neither  on  the  design  of  the  experiments  nor  on  their  results,  but  rather 
on  the  relation  of  the  methods  used  to  the  theory. 


A.  Current  Status  of  Positive  Evidence 

The  major  body  of  positive  evidence  for  the  theory  lies  in  the  field 
of  accumulated  clinical  observations.  The  first  achievement  of  the  system 
was  a  phenomenological  one:  it  called  attention  to  a  vast  array  of 
phenomena  and  to  the  relations  between  them,  and  for  the  first  time 
made  these  appear  meaningful  and  amenable  to  rational  consideration. 
In  regard  to  these  phenomena  and  relationships,  the  accumulated  clinical 
evidence  is  positive  and  decisive.  The  situation  is  different,  however,  in 
regard  to  the  theoretical  propositions  of  the  system.  While  the  evidence 

90  See  the  introductions  to  these  volumes  and  their  other  references  to  Freud. 

The  Structure  of  Psychoanalytic  Theory  141 

In  respect  to  these  also  seems  massive  and  Imposing,  the  lack  of  clarifica- 
tion as  to  what  constitutes  a  valid  clinical  research  method  leaves  unde- 
termined the  positive  evidential  weight  of  the  confirming  clinical  material 
In  spite  of  the  various  discussions  [e.g.,  Brenman  and  Gill,  34,  147, 
Kubie,  209,  210;  Benjamin,  11;  Escalona,  67,  etc.]  on  the  nature  oi 
clinical  research,  and  in  spite  of  French's  [89]  extensive  attempt  tc 
exemplify  the  method,  its  principles  have  not  yet  been  expressed  in  the 
form  of  a  canon.  Indeed,  many  psychologists  would  question  whethei 
there  is  or  can  be  any  other  canon  of  research  than  the  experimental 
Since  it  is  questionable  whether  there  exists  such  a  thing  as  the  experi- 
mental canon,  these  views  need  not  worn-  us.  Because  a  canon  of  clinical 
investigation  is  lacking,  much  of  the  evidence  for  the  theory  remains 
phenornenological  and  anecdotal,  even  if  its  obviousness  and  bulk  tend  tc 
lend  it  a  semblance  of  objective  validity.  This  makes  it  urgent  to  rein- 
vestigate  Freud's  case  studies  with  the  aim  of  clarifying  whether  or  not 
they  can  yield  a  canon  of  clinical  research  at  the  present  stage  of  oui 

In  the  lack  of  a  canon  for  clinical  research,  it  is  difficult  to  accept  as 
positive  evidence  observations  which  must  first  be  interpreted  before  it 
becomes  clear  whether  or  not  they  confirm  the  predictions  of  the  theory : 
we  must  be  wary  lest  we  smuggle  in  the  confirmation  through  the  inter- 
pretation. Axiomatization  and/or  a  canon  of  investigation  protect  other 
sciences  from  such  circularity.  The  lack  of  such  safeguards  is  a  real 
handicap  for  this  theory,  since  by  the  very  nature  of  the  relation  between 
observations  and  theory,  only  observations  pertaining  to  basic  concepts 
and  theorems  can  be  free  of  interpretation  (cf.  pp.  116-121).  For  in- 
stance, one  of  the  major  propositions  of  the  psychoanalytic  theory,  con- 
firmed by  observations,  is  that  there  are  two  kinds  of  mental  processes: 
primary  and  secondary.  Little  or  no  interpretation  of  the  observations  is 
needed  to  demonstrate  that  pathological,  dream,  or  drug  states  bring  tc 
the  fore  mental  processes  which  do  not  abide  by  the  laws  of  orderec 
logical  thought.  But  only  on  this  low  level  of  abstraction  is  the  evidence 
conclusive  without  interpretation.  As  soon  as  the  evidence  for  the 
mechanisms  of  the  primary  process  is  tackled,  observation  and  interpreta- 
tion begin  to  shade  into  each  other.  Per  se,  that  should  not  invalidate  th< 
evidence,  since  no  science  can  get  along  without  interpreting  its  findings 
Yet  in  psychoanalysis  the  difficulty  is  that  the  canon  of  interpretatioi 
itself  is  in  question — or  at  least  not  beyond  question — and  it  is  likely  t< 
remain  so  until  the  nature  of  the  clinical  method  has  been  clarified,  o 
until  experimental  methods  have  been  found  which  provide  an  inde 
pendent  base  for  the  theory.  As  things  stand,  there  is  no  canon  whereb; 
valid  interpretation  can  be  distinguished  from  speculation,  though  0, 
post  facto  the  experienced  clinician  can  distinguish  them  rather  well. 


We  have  a  few  experiments  which  are  free  of  this  difficulty.  The  ex- 
periments on  dream  symbolism  [306,  288]  and  the  related  observations 
of  Silberer  [310;  see  also  Rapaport,  276]  are  the  outstanding  ones.  But 
these  experiments  remain  phenomenological  in  that  they  demonstrate  the 
existence  of  symbolization,  rather  than  the  specific  conditions  of  its 
occurrence.  The  Poetzl  [257]  experiment  and  Fisher's  [84]  replication  of 
it,  as  impressive  as  they  are,  involve  interpretation.91 

Most  of  the  experimental  evidence  for  the  theory  is  questionable, 
even  if  Sears5  survey  [308],  which  was  loaded  with  negative  bias,  was  in- 
clined to  accept  some  of  it  as  positive  evidence.  The  overwhelming  ma- 
jority of  experiments  designed  to  test  psychoanalytic  propositions  display 
a  blatant  lack  of  interest  in  the  meaning,  within  the  theory  of  psycho- 
analysis, of  the  propositions  tested.  Thus  most  of  them  certainly  did  not 
measure  what  they  purported  to;  as  for  the  rest,  it  is  unclear  whether  or 
not  they  did.  Even  where  the  findings  appear  to  confirm  a  relationship 
posited  by  psychoanalysis,  the  experiments  usually  tested  only  an  anal- 
ogous relationship  on  a  high  level  of  the  hierarchy  of  psychological  or- 
ganization. It  is  not  that  all  these  experiments  are  useless  as  confirming 
evidence,  but  rather  that  at  this  stage  of  our  knowledge  it  is  not  clear 
what — if  anything — they  confirm.  It  is  hard  to  share  Hilgard's  [173] 
enthusiasm  for  most  of  the  experiments  he  considers  relevant  and  con- 
firming. It  is  likely  that  some  of  the  experimental  findings  will  fall  into 
place  when  ego  psychology  has  clarified  the  hierarchic  relationships 
which  obtain  in  psychological  organization.  Command  of  the  theory 
should  help  toward  making  the  results  of  future  experiments  unequivocal, 
but  it  is  not  as  much  of  a  guarantee  of  success  as  ignorance  of  the  theory 
is  of  failure.  The  experimental  psychologist  who  enters  the  precincts  of 
psychodynamics  meets  the  same  complexities  which  the  clinical  ob- 
server has  been  struggling  with  for  over  six  decades.  There  are  no  "easy 
pickings"  and  the  "experimental  method"  has  no  magic  here. 

In  conclusion:  the  extensive  experimental  evidence  for  the  system, 
which  would  seem  to  confirm  it  in  terms  of  the  usual  criteria  of  psycho- 
logical experiments,  cannot  be  considered  conclusive  in  terms  of  the 
psychoanalytic  theory,  since  most  of  the  experiments  disregard  the  the- 
ory's definitions.  The  extensive  clinical  evidence,  which  would  seem  con- 
clusive in  terms  of  the  system's  internal  consistency,  fails  to  be  con- 
clusive in  terms  of  the  usual  criteria  of  science,  because  there  is  no  estab- 

91 1  am  not  listing  here  D.  Levy's  and  J.  McV.  Hunt's  experiments,  nor  others 
akin  to  them,  because  they  are  animal  experiments,  and  represent  conditions  of  a 
simplicity  which  does  not  obtain  in  man. — By  the  time  of  proofreading  I  had  an 
opportunity  to  read  the  manuscript  of  the  joint  study  by  C.  Fisher  and  I.  Paul 
presented  at  the  1958  meetings  of  the  American  Psychological  Association.  It  goes 
a  long  way  toward  meeting  the  difficulties  discussed  here. 

The  Structure  of  Psychoanalytic  Theory  143 

lished  canon  for  the  interpretation^  of  clinical  observations.  Thus,  only 
a  few  observations  and  experiments  (themselves  in  need  of  replication) 
offer  evidence  acceptable  both  in  terms  of  the  theory  and  in  terms  of 
psychology  at  large. 

B.  Major  Sources  of  Incompatible  Data 

It  is  often  assumed  that  the  data  and  theories  of  the  "dissident" 
schools  of  psychoanalysis  [cf .  Munroe,  240]  are  incompatible  with  and  an 
embarrassment  to  psychoanalytic  theory.  This  does  not  seem  to  be  the 
case.  Recent  developments  [for  instance,  Zetzel,  330,  Erikson,  66]  in 
psychoanalytic  theory  in  general,  and  in  ego  psychology  in  particular, 
seem  to  show  that  this  theory  has  the  foundations  for  concepts  and  propo- 
sitions which  can  account  for  the  observations  made  and  the  valid  rela- 
tionships posited  by  the  dissident  schools.  Thus  we  find  no  source  of 
embarrassment  here,  but  rather  a  task  to  be  accomplished. 

It  is  at  times  assumed — particularly  by  psychologists — that  the  find- 
ings and  therapeutic  results  of  Rogers'  client-centered  therapy  are  a 
source  of  embarrassment  for  the  psychoanalytic  theory.  But  this  is  hardly 
the  case.  Rogers3  counseling  procedure,  at  least  to  begin  with,  had  no 
general  psychological  theory,  nor  even  a  theory  of  personality.  The  vague 
outlines  of  a  theory  of  personality,  which  it  has  developed  since  then, 
seem  to  form  a  segment  of  an  ego  psychology.  Thus  the  possibility  of 
contradiction  and  embarrassment  is  limited  to  begin  with,  and  is  further 
minimized  by  two  other  aspects  of  nondirective  counseling.  First,  the  very 
idea  of  nondirectiveness  is  one  of  the  implications  of  psychoanalytic 
therapy.  The  method  of  free  association  and  the  analyst's  "evenly  hover- 
ing" [106]  attention  imply  it.  They  both  demand  that  the  patient's  prob- 
lems not  be  prejudged  and  that  reliance  be  placed  on  his  ability  to  meet 
his  problems  spontaneously.  Rogers'  criticism  of  psychoanalysis  is  well 
founded  in  so  far  as  it  implies  that  with  the  accumulation  of  psycho- 
analytic knowledge,  and  with  experience  in  wielding  the  tools  of  inter- 
pretation, often  little  room  was  left  for  the  patient's  spontaneity,  and  too 
often  the  therapist  came  to  be  always  right  and  the  patient  always  wrong. 
Indeed,  psychoanalytic  ego  psychology,  too,  may  be  considered  to  be  a 
reaction  to  this  danger,  and  the  emphasis  on  the  activity  of  the  ego  as  a 
crux  of  therapy  seems  to  have  a  central  place  in  its  therapeutic  and  gen- 
eral theory.93  But  the  roots  of  this  danger  are  in  the  practice  rather  than 
in  the  theory  even  of  "classical"  psychoanalysis.  Second,  nondirectiveness 
is  but  one  aspect  of  the  technique  of  psychoanalysis,  and  can  be  no  more 

82  The  nonexistent  scientific  canon  of  interpretation  is  not  to  be  mistaken  for 
the  well-established  clinical  techniques  of  interpretation. 

S3Cf.  Sullivan,  Homey;  also  P.  Bergman  [12],  Gill  [145],  and  Rapaport  [266, 
275,  280]. 


than  one  aspect  of  any  other  therapy.  Experience  has  confronted  non- 
directive  counselors  with  the  problems  of  transference  and  resistance 
familiar  to  open-eyed  therapists  of  any  persuasion.  When  the  "nondirec- 
tive  approach"  faces  these  problems,  it  will  meet  the  eternal  struggle  of 
man's  spontaneity,  goodness,  readiness  and  ability  to  help  himself,  against 
man's  inertia,  fear  of  his  own  spontaneity,  need  for  help,  etc.  While  it 
is  true  that  treating  man  as  a  helpless,  inert,  and  needful  creature  is  prone 
to  demobilize  his  spontaneity  and  ability  to  help  himself,  it  is  also  true 
that  man's  helplessness,  inertia,  and  need  for  support  will  not  be  elim- 
inated by  denying  that  they  exist.  Therapies  or  therapists  who  practice 
either  sort  of  denial  end  up  by  establishing  their  own  McCarran  Act: 
sooner  or  later  they  announce  that  this  or  that  kind  of  patient  is  not 
the  right  kind  for  their  kind  of  therapy.  Not  rarely  they  go  further 
and  announce  that  this  or  that  kind  of  patient  is  "not  treatable."  In  the 
long  run,  psychological  theories  of  therapy  must  come  to  a  point  where 
they  will  make  it  possible  to  select  the  therapy  which  is  good  for  a  pa- 
tient and  not  the  patient  who  is  good  for  a  therapy.94  Yet  Rogers'  suc- 
cesses, limited  though  they  may  be,  clearly  show  how  little  we  know 
about  the  ego,  its  activity  and  passivity,  its  sources  of  energy,  etc.  Reider's 
[285]  report  of  "spontaneous  cures"  likewise  shows  up  our  ignorance.  In 
this  sense,  though  not  "embarrassing"  or  "inconsistent,"  Rogers  seems  to 
provide  data  which  prod  psychoanalysis  toward  further  exploration  of 
familiar  as  well  as  barely  charted  areas  of  ego  psychology. 

Many  psychologists  and  even  psychoanalysts  (particularly,  but  not 
only,  Neo-Freudians)  have  assumed  that  Dollard  and  Miller's  [48]  study 
and  Mowrer's  [239]  experiments  and  their  theoretical  combination  of 
psychoanalysis  with  learning  theory  have  cut  across  the  "theoretical 
jungle"  of  psychoanalysis,  replacing  much  of  it  by  learning  theory.  The 
powerful  position  occupied  by  learning  theory  until  recently  (and  perhaps 
still)  on  the  American  scene  "reinforced"  this  assumption.  But  the  fate  of 
psychoanalytic  theory,  or  for  that  matter,  the  fate  of  any  theory,  cannot 
be  settled  by  popular  vote;  if  it  could  be,  psychoanalysis  would  be  in  a 
bad  way.  Learning  theory  seems  to  be  the  (academic)  theoretical  back- 
bone of  the  majority  of  recent,  mass-produced  clinical  psychologists.  But 
since  this  theory  cannot  guide  their  clinical  work,  they  rely  there  in- 
creasingly upon  psychoanalytic  propositions,  whose  theory  they  have  not 
studied.  Thus  the  "marriage  of  convenience"  that  Bollard,  Miller,  and 
Mowrer  recommend  between  psychoanalysis  and  learning  theory  must 
indeed  seem  to  be  convenient  to  them,  since  it  seems  to  justify  clin- 
ical practices,  while  at  the  same  time  it  provides  a  salve  to  the  aca- 
demic conscience.  Does  the  work  of  these  authors  provide  data  in- 

94  See  Knight  [202,  203]  and  Gill  [145]  concerning  the  bearing  of  psycho- 
analytic ego  psychology  on  these  issues. 

The  Structure  of  Psychoanalytic  Theory  14!! 

consistent  with  and  embarrassing  to  psychoanalytic  theory?  Since  the] 
present  themselves  primarily  as  protagonists  of  psychoanalysis  and  pur- 
port to  provide  it  with  a  solid  experimental  and  conceptual  foundation 
this  question  is  not  easily  answered.  Still,  we  must  ask:  while  these 
authors  (excepting  in  some  respects  Mowrer)  do  not  intend  to  embarrass 
psychoanalysis,  have  they  nevertheless  produced  data  incompatible  witib 
psychoanalytic  findings  and  theories?  Only  a  brief  sketch  of  the  theo- 
retical situation  can  be  attempted  here  [see  Rapaport,  271,  2721. 

These  investigators  have  produced,  by  the  method  of  conditioning, 
experimental  analogues  of  "Freudian  mechanisms"  in  animals  [Masser- 
man,  229].  These  analogues  would  be  neither  embarrassing  to  nor  in- 
compatible with  psychoanalytic  theory7  if  no  claim  were  made  that  in 
man,  too,  the  mechanisms  of  the  primary  process  and  of  the  defenses 
are  products  of  conditioning.  Bollard,  Miller,  Mowrer,  and  Masserman 
imply — to  say  the  least — such  a  claim,  and  thereby  elevate  the  condition- 
ing theory  of  learning  to  the  status  of  the  learning  theory  of  psycho- 
analysis. This  is  incompatible  with  psychoanalytic  theory,  since  it  makes 
the  economic  and  genetic  points  of  view  superfluous  and  thus  clashes 
with  the  observational  data  which  made  these  points  of  view  necessary1 
parts  of  the  theory  (cf.  pp.  110-114).  Psychoanalytic  theory  at  present 
cannot  escape  this  embarrassment,  since  it  has  no  learning  theory  of  its 
own  to  pit  against  conditioning.  This  lack  is  not  palliated  by  the  demon- 
stration that  the  conditioning  theory  of  learning  does  not  meet  the 
empirical  requirements  (e.g.,  automatization  problems,  structure  forma- 
tion, distinction  between  primary  and  secondary  processes)  which  a  psy- 
choanalytic learning  theory  will  have  to  meet.  Psychoanalysis  will  be 
totally  free  of  embarrassment  from  this  quarter  only  when  it  has  a  learn- 
ing theory  which  not  only  fulfills  its  own  empirical  and  theoretical  re- 
quirements, but  is  also  broad  enough  to  account  for  conditioning 
phenomena — including  the  conditioned  analogues  of  "unconscious  me- 
chanisms"— as  special  cases. 

The  work  of  these  investigators  has  come  up  against  the  problem  of 
persisting  psychic  formations,  which  has  beset  and  embarrassed  all 
motivational  (need-gratification)  theories.  G.  Allport's  criticism  of 
motivational  theories  and  his  ego  psychology  start  from  this  problem,  and 
psychoanalytic  ego  psychology  faces  it  squarely.  The  method  of  condition- 
ing used  by  Miller,  Dollard,  etc.,  determined  the  form  in  which  they  en- 
countered this  problem:  conditioned  responses  are  in  general  subjed 
to  extinction ;  thus,  abiding  psychological  formations  require  explanation. 
Why  are  they  exempt  from  this  rule?  or,  How  are  they  so  reinforced  as 
to  avoid  extinction?  This  is  indeed  one  of  the  central  difficulties  of  al 
conditioning  theories  of  learning  [cf.  82].  The  theory  of  neuroses  brings 
these  questions  into  sharp  relief,  since  symptoms  are  apparently  non- 


rewarding  and  should  thus  be  subject  to  extinction.95  Bollard  and  Miller 
as  well  as  Mowrer  tried  to  meet  the  problem  by  assuming  that  reinforce- 
ment through  "learned"  (conditioned)  drives  can  account  for  non- 
extinction.  This  solution  brings  with  it  the  same  difficulties  which  raised 
the  problem  to  begin  with — namely,  that  drives,  whether  learned  or 
not,  cannot  account  for  persisting  structures.  But  this  fact  does  not  seem 
to  have  deterred  Bollard  and  Miller,  though  they  were  aware  that  the 
problem  is  one  of  ego  psychology. 

Mowrer,  however,  was  apparently  not  satisfied  with  postulating — 
and  demonstrating  by  analogues — "acquired  drives,'3  but  asserted  that 
they  are  acquired  by  contiguity  and  not  by  reinforcement  learning.  This 
assertion  and  the  observations  it  is  based  on — though  questioned  by 
learning  theorists — are  a  source  of  embarrassment  for  psychoanalytic 
theory,  and  will  continue  to  be  as  long  as  psychoanalytic  theory 
accounts  for  derivative  drives  by  differentiation  of  basic  drives,  and  infers 
that  this  occurs  parallel  with  structure  development,  but  cannot  specify 
either  the  process  of  structure  formation  or  that  of  drive  differentiation. 

But  what  is  more  often  considered  a  source  of  embarrassment  for  psy- 
choanalytic theory  in  Mowrer's  system — his  formulation  that  neurosis  is 
due  to  "underlearning"  and  not  to  "overlearning" — is  actually  no  source 
of  inconsistent  data.  Mowrer  apparently  saw  that  a  conditioning  theory 
(whether  monistic  or  dualistic)  can  hardly  explain  the  persistence  of 
"learned  drives"  and  nonrewarded  symptoms  by  "overlearning."  There- 
fore, he  reasoned,  if  the  drives  and  the  neurotic  drive  manifestations  can- 
not be  proved  to  be  "overlearned,"  then  that  which  is  supposed  to  control 
them  must  be  "underlearned."  Thus  he  equated  the  repressing  forces 
(censorship,  superego)  which — according  to  him — are  weak  in  neuroses, 
with  underlearned  social  prohibitions.  This  sounds  logical,  but  it  is  not 
psychological,  and  is  doubly  incompatible  with  psychoanalytic  theory 
and  observations.  First,  it  implies  what  is  to  be  proved,  namely  that  the 
intrapsychic  structures  and  forces  in  question  are  learned  (conditioned). 
Second,  it  implies  that  these  structures  and  forces  are  ineffective  because 
of  their  weakness  or  absence  ("underlearning"),  though  the  concept  of 
the  unconscious  in  general,  and  the  observations  concerning  the  un- 
conscious sense  of  guilt  in  particular  provide  a  different  explanation  so 
far  not  contested  by  any  evidence.  Having  replaced  the  "overlearning 
theory"  which,  according  to  him,  is  the  core  of  the  psychoanalytic  "drive- 
repression  theory"  of  neurosis,  he  assumes  that  he  has  demolished  the 
latter.  His  "underlearning  theory"  of  neurosis,  translated  into  clinical 
language,  says  that  the  trouble  with  neurotics  is  not  that  their  censorship 
(repressive  forces,  conscience)  is  too  strong,  but  rather  that  it  is  too 
weak,  having  been  repressed  by  the  id  and  ego  combination.  However, 

95  N.  Maier  [228]  took  this  issue  so  to  heart  that  he  propounded  a  dualistic 
theory  of  motivation  learning  vs.  frustration  learning. 

The  Structure  of  Psychoanalytic  Theory  147 

he  does  not  specify  the  process  and  mechanism  of  this  "repression."  For 
Mowrer  the  neurotic  is  a  person  who  "did  not  learn  the  lesson.35  He  takes 
us  back  to  moralistic  and  religious  precepts  and  to  the  pre-Freudian 
conception  of  neurosis.  What  seems  to  have  happened  is  that  Mowrer 
rediscovered  the  unconscious  sense  of  guilt,  long  since  recognized  in 
psychoanalysis.  Not  realizing  the  place  of  his  "discovery55  in  psycho- 
analytic theory,  he  explained  it  in  terms  of  learning  theory  and  put  it  in 
the  center  of  psychodynamics,  unconcerned  with  the  consequences  of 
this  recentering  for  psychodynamics  at  large. 

Piaget's  observations  and  theories,  if  confirmed,  may — though  they 
need  not — prove  to  be  sources  of  incompatible  data.  They  seem  to 
demonstrate  that  structure  (schema)  formation  arises  from  disturbances 
in  the  equilibrium  of  existing  structures  (schemata),  and  that  such 
disturbances  always  act  as  motivations  (desirability).  In  Piaget's  terms, 
function  always  brings  about  structural  change  (disequilibrium)  which, 
in  turn,  provides  motivation  (desirability)  for  a  repetition  of  the  function 
(circular  reaction)  which  consolidates  the  structural  change,  i.e.,  builds 
new  structure  (schema) .  Now  it  may  prove  possible  to  treat  the  observa- 
tions on  which  this  theory  is  based  in  terms  of  what  psychoanalysis  calls 
processes  of  autonomous  ego  development.  If  so,  Piaget's  theory  would 
shed  new  light  on  the  nature  of  many  ego  motivations  and  would  cor- 
roborate Hartmann's  assumption  that  the  ego  has  sources  of  energy  other 
than  bound  and  neutralized  drive  cathexes.  It  would  also  force  us  to 
rethink  the  theory  of  id-ego  relationships.  But  if  Piaget's  theory  and  the 
observations  it  is  based  on  should  lead  to  the  conclusion  that  the  only 
source  of  motivations  is  the  one  discovered  by  him,  they  would  become 
incompatible  with  psychoanalytic  theory.  In  either  case  nothing  more 
desirable  could  happen  to  psychoanalytic  theory  than  a  corroboration 
of  Piaget's  findings.  Psychoanalysis  would  find  itself  for  the  first  time 
confronted  with  a  genetic  theory  of  broad  scope,  using  a  method  of 
observation  which  is  in  some  ways  akin  to  (if  not  derived  from)  its  own. 
The  mutual  stimulation  of  this  confrontation  could  not  but  prove 

Psychologists,  particularly  experimental  psychologists,  seem  to  assume 
that  experimental  tests  of  psychoanalytic  theories,  if  negative  in  outcome, 
provide  data  inconsistent  with  and  embarrassing  to  the  theory.  Sears 
[308]  and  the  many  who  have  quoted  him  and  relied  on  him  seem  to 
have  assumed  something  like  this.  Would  that  it  were  so.  It  is  not.  Most 
of  the  studies  Sears  surveyed  took  a  psychoanalytic  statement  out  of  its 

^Harlow's,  Christie's,  and  others'  observations  concerning  "activity  drives" 
may  also  pose  a  problem  akin  to  that  posed  by  Piaget's  studies.  But  these  observa- 
tions are  too  new  to  be  assessed. 


context  and  tested  the  statement,  rather  than  the  theory,  which  they 
usually  knew  little  about.  Moreover,  they  used  methods  of  testing  alien 
to  the  observations  from  which  the  statement  and  its  terms  derived.  It  is 
doubtful  that  any  of  the  currently  available  experimental  results  can  be 
proved  clearly  incompatible  with  the  theory.  Here  the  very  difficulty  of 
obtaining  data  inconsistent  with  and  embarrassing  to  the  system  becomes 
an  embarrassment  to  it.  Psychoanalytic  theory,  which  is  adequate  for 
clinical  purposes,  will  have  to  become  much  more  systematic  before  ex- 
periments can  be  designed  which  will  not  simply  confirm  or  refute  its 
propositions,  but  rather  specify  and  modify  them.  Thus  the  experimental 
psychologist  who  approaches  it  must  assume  the  responsibility  of  clarify- 
ing and  specifying  theoretically  the  propositions  which  he  undertakes  to 
test.  For  the  time  being  this  is  the  only  way  to  arrive  at  experimental 
findings  relevant  to  and  incompatible  with  the  theory. 

C.  "Critical"  Tests  of  Principal  Assumptions 

Those  difficulties  in  testing  psychoanalytic  propositions  which  we 
have  discussed  naturally  apply  also  to  the  so-called  critical  tests.  In 
addition,  the  latter  usually  require  the  existence  of  alternative  theories  or 
alternative  possibilities  within  the  theory.  There  are  few,  if  any,  specific 
psychoanalytic  propositions  for  which  other  theories  have  an  alternative 
to  offer,  and  since  the  psychoanalytic  theory  itself  is  not  geared  to  ex- 
perimental tests,  it  does  not  usually  envisage  alternatives  in  the  sense  im- 
plied by  the  conception  of  crucial  tests,  but  rather  in  that  implied  by 
alternative  interpretations.  While  the  alternatives  in  the  former  sense  call 
for  a  decision  between  two  possibilities,  one  of  which  is  incompatible  with 
the  theory,  the  alternatives  in  the  latter  sense  are  both  consistent  with  the 
theory,  but  only  one  is  realized  in  the  phenomenon  studied,  whereas  the 
other  is  not.  The  former  pertains  to  systematic  possibilities,  the  latter 
either  to  a  single  instance,  or  to  a  specific  genetic  sequence,  or  to  an 
individual  person.  Thus  critical  tests  are  hardly  possible  for  the  proposi- 
tions of  the  special  (clinical)  theory  of  psychoanalysis.97  The  opportunities 
— if  any — for  such  tests  must,  then,  be  sought  in  the  general  (psy- 
chological) theory  of  psychoanalysis.  But  the  primitive  state  of  the 
systematization  of  this  general  theory  militates  against  the  possibility  of 
critical  tests,  and  so  does  the  nonexistence  of  other  theories  of  com- 
parable scope. 

For  these  reasons  it  is  difficult  to  envisage  "critical"  tests  of  this 

97  Clinical  predictions  are  always  fraught  with  the  fact  that  all  motivations  have 
multiple,  equivalent,  alternative  means  and  goals.  Thus,  such  predictions  usually 
cannot  specify  which  of  these  equivalent  alternatives  are  to  be  expected,  and 
therefore,  the  results  of  experimental  tests  of  predictions  must  first  be  interpreted 
before  their  bearing  on  the  theory  can  be  established. 

The  Structure  of  Psychoanalytic  Theory  149 

theory.  Yet  it  is  probable  that  the  sources  of  actually  or  potentially  in- 
compatible data — i.e.,  learning  theories  and  Piagefs  work — must  serve 
as  points  of  departure  for  critical  tests. 

Critical  studies  centering  around  Piaget's  theory  must  first  cor- 
roborate his  observations  and  extend  them  to  behavior  which  involves 
affects  and  motivations,  in  the  psychoanalytic  sense  of  these  terms.  The 
aim  of  studies  revolving  around  learning  theory  would  be  to  demonstrate 
processes  of  structure  building  and  learning  compatible  with  psycho- 
analytic theory,  but  incompatible  with  existing  learning  theories,  or  vice 
versa.  Any  quantitative  method  may  lead  to  a  critical  test  if  it  can  trace 
qualitatively  the  process  of  structure  consolidation,  that  is  to  say,  if  it 
can  show  that,  once  a  certain  set  of  qualitative  changes  has  occurred  in 
a  process  of  acquisition,  a  nonextinguishing  structure  arises.  It  will  turn 
into  a  critical  test  when  it  can  also  show  that  existing  learning  theories 
are  incompatible  with  the  process  of  structure  formation  traced  by  it. 
None  of  the  well-known  methods  can  at  present  be  regarded  as  the  royal 
road  to  such  a  critical  test.  But  the  following  may  serve  as  examples 
of  techniques  which  might  be  tried:  tracing  quantitatively  the  qualita- 
tive changes  in  the  acquisition  of  skills  which  are  not  simply  compounds 
of  other  skills;  tracing  the  qualitative  changes  in  the  course  of  learning 
meaningful  verbal  material;  tracing  how  subjects  spontaneously  discover 
a  meaning  or  a  pattern  embedded  in  material  wrhich  they  handle  in  the 
deliberate  pursuit  of  a  different  goal.  Altogether,  any  quantitative  tech- 
nique which  makes  it  possible  to  follow  the  qualitative  (and  not  just  the 
quantitative)  course  of  the  development  of  any  behavior,  which  is  on  its 
way  to  becoming  a  part  of  the  person's  quasi-abiding  behavior  equip- 
ment, might  conceivably  become  the  method  of  choice  for  a  critical  test. 

This  lengthy  discussion  of  "critical"  tests  is  warranted  neither  by  the 
actual  state  of  psychoanalytic  theory  nor  by  my  knowledge  of  these 
matters.  Its  purpose  is  to  stress  that  crucial  tests — if  they  are  to  come — 
will  not  necessarily  center  on  motivations.  Indeed,  my  intention  here 
has  been  to  make  it  plausible  that  the  crucial  experimental  contribution 
toward  the  consolidation  of  psychoanalytic  theory  may  well  be  made  at 
an  apparent  distance  from  what  is  commonly  considered  its  home  ground. 
It  may  well  come  on  the  battlefield  of  learning  theory,  or  on  that  of 


A.  The  Range  of  Application 

Unlike  most  psychological  theories,  whose  application  outside  their 
initial  ground  is  a  matter  of  future  possibility  or  probability,  the  applica- 


tion  of  psychoanalysis  to  nearly  all  human  endeavors  and  products  has 
been  envisaged  and  actualized  from  the  very  beginning. 

The  questions  here  are  not  what  applications  are  possible,  but 
rather,  how  valid  and  effective  are  they?  How  can  they  be  made  more 
appropriate  and  effective?  Since  these  applications  have  rarely  been 
systematic,  the  need  for  and  the  possibility  of  studies  aimed  at  systematiz- 
ing the  existing  applications  are  practically  unlimited.  Such  studies  might 
well  increase  the  effectiveness  of  the  applications  by  bringing  them  in 
line  with  the  present  state  of  development  and  systematization  of  psy- 
choanalytic theory.  The  development  of  ego  psychology  provided  psycho- 
analysis with  new  tools  which  bid  fair  to  increase  the  appropriateness  of 
its  application  in  all  fields.  Particularly  the  application  of  psychoanalysis 
to  sociology*  and  anthropology  has  gained  and  stands  to  gain  further 
from  Hartmann's  [159]  and  Erikson's  [60,  61,  66]  contributions,  like- 
wise, its  applications  to  art  from  the  contributions  of  Kris  [207].  The 
change  that  ego  psychology  has  wrought  in  the  relationship  between  psy- 
choanalysis and  psychology  has  been  discussed  above. 

B.  Methods.,  Concepts,  and  Principles  of  Long-term  Significance 

In  his  outline,  Dr.  Koch  defines  "long-term  significance"  as  the 
ability  to  survive  independently  from  "the  over-all  structure  or  detailed 
assumptional  content  of  the  system."  Freud  repeatedly  stated  that  an) 
therapy  which  takes  into  account  the  unconscious,  transference,  and 
resistances  is  psychoanalysis.  Thus  the  concepts  which  he  considered  tc 
be  of  broadest  significance  are  the  dynamic  unconscious,  transference 
and  resistance. 

But  perhaps  we  can  go  beyond  Freud's  view  if  we  consider  first 
that  the  methods,  principles,  and  concepts  of  greatest  independence  ii 
any  system  are  on  the  one  hand  those  closest  to  observations  and,  on  th 
other,  those  of  greatest  generality;  second,  that  some  of  the  methods- 
principles,  and  concepts  of  all  major  theories  sooner  or  later  become  s< 
general  that  they  enter  the  public  domain  and  can  no  longer  be  COD 
sidered  specific  to  the  theory.  Psychoanalysis  has  developed  method: 
concepts,  and  principles  which  are  now  in  the  public  domain:  for  e? 
ample,  the  method  of  interview;98  the  concepts  of  the  "descriptive  ur 
conscious/'  motivation,  and  defense;  and  the  principle  of  psychologic; 
determinism.  It  could  be  justly  argued  that  all  of  these  antedate  ps^ 
choanalysis.  But  psychoanalysis  changed  their  character  and  gained  fc 
them  an  acceptance  in  the  public  domain,  where  they  are  now  ii 
dependent  of  the  theory  and  not  subject  to  its  changes. 

method  of  interview  implies  only  that  the  past  is  relevant  to  the  unde 
standing  of  the  present,  but  no  other  specifically  psychoanalytic  assumption 
concept  [cf.  166]. 

The  Structure  of  Psychoanalytic  Theory  151 

1.  Methods.  What  are  the  methods  of  psychoanalysis?  It  may  be  use- 
ful to  distinguish  here  between  methods  and  techniques.  Let  us  define 
techniques  as  the  specific  ways  and  means  by  which  methods  are 
applied  and  note  that  in  psychoanalysis  they  have  the  additional  con- 
notation of  ways  and  means  which  are  not  only  exploratory,  but  also 
effective  therapeutically.  The  techniques  of  psychoanalysis  have  been 
studied  [106,  72,  151]  but  its  methods  have  scarcely  been  given  system- 
atic thought  [see,  however,  Bemfeld,  14].  Thus,  what  follows  can  be  no 
more  than  a  preliminary7  sketch. 

It  would  seem  that  the  basic  method  of  psychoanalysis"  is  the 
method  of  interpersonal  relation  [314,  260];  more  specifically,  it  is  the 
participant  observation  variant  of  the  method  of  interpersonal  relation; 
in  particular,  it  applies  the  nondirective  (free  association),  the  interpre- 
tive-genetic, and  the  defense-analysis  techniques  of  participant  observa- 
tion [see  Gross,  153;  also  Rapaport,  263,  273].  These  methods  and  tech- 
niques, unlike  the  interview  method,  are  linked  to  the  theory  of  psycho- 
analysis in  that  the  phenomena  they  are  based  on  are  the  observational 
referents  of  the  transference  concept.  Human  beings  in  dealing  with  each 
other  repeat  the  patterns  they  have  developed  in  their  relations  to 
"significant  others,"  and  these  patterns  of  relationships  ultimately  go 
back  to  those  which  the  individual  has  developed  toward  the  earliest 
"significant  others" :  father,  mother,  siblings,  nurses,  etc.  Such  repetitions 
of  relationship  patterns  are  the  empirical  referents  of  the  transference 
concept.  Transferences  are  ubiquitous  in  everyday  life,  but  so  far  the  psy- 
choanalytic methods  are  the  only  ones  for  observing  them  systematically 
and  for  tracing  their  genetic  roots.  The  aim  of  the  psychoanalytic  method 
of  interpersonal  relation  is  to  bring  about  such  transferences.  The  aim 
of  the  method  of  participant  observation  is  to  make  these  transferred 
patterns  conscious.  The  free  association,  interpretive-genetic,  and  defense- 
analysis  techniques  are  specific  interventions  facilitating  insight  into  these 

In  so  far  as  these  methods  and  techniques  are  tied  to  the  concept  of 
transference  they  are  specifically  psychoanalytic.100  But  they  are  so  closely 
related  to  a  broad  and  crucial  range  of  observations  that  it  is  hard  to 
conceive  of  changes  in  the  structure  of  psychoanalytic  theory  which  would 
alter  them  or  dispense  with  them.  What  has  changed  repeatedly,  and  is 
likely  to  change  again,  is  the  relative  emphasis  in  the  theory  and  in 
practice  on  any  one  of  these  methods  and  on  the  patient's  gaining  in- 
sight into  his  transference  patterns.  Interview  and  therapy  methods  which 

99  Here  we  are  concerned  only  with  the  methods  specific  to  psychoanalysis  and 
disregard  others  like  suggestion,  support,  etc.  [cf.  Bibring,  28]. 

100  They  vary  in  this  respect;  of  the  three,  the  free-association  technique  seems 
to  be  the  one  least  closely  tied  to  the  concept  of  transference. 


do  not  aim  at  gaining  information  about  and  insight  into  transference 
patterns  may  well  achieve  their  limited  or  different  goals,  but  none  so 
far  has  succeeded  in  replacing  the  psychoanalytic  methods  of  system- 
atically observing  transference  patterns.  Projective  techniques  do  obtain 
some  such  data,  but  the  recent  emphasis  on  the  significance  of  the  inter- 
personal relation  between  patient  and  tester  [see  Schafer,  298]  points  to 
their  limitations.  Whatever  the  fate  of  those  more  specific  methods 
described  as  "techniques,"  and  whatever  the  ultimate  judgment  on  the 
therapeutic  effectiveness  of  these  basic  psychoanalytic  methods,  the  latter 
are  likely  to  stay  with  us  as  unique  methods  of  observation  for  a  very 
long  while. 

2.  Principles.  The  "points  of  view"  seem  to  be  the  equivalents  of 
"principles"  in  psychoanalytic  theory.  Yet  their  form  shows  that  the 
time  to  examine  them  one  by  one,  for  their  long-range  significance,  has 
not  yet  arrived.  Instead  of  formal  principles  we  will  present  here  a  fewT 
general  conceptions,  which  compound  the  various  points  of  view,  and 
which  seem  likely  to  survive  whatever  the  fate  of  the  more  specific 
ingredients  of  the  psychoanalytic  theory  should  prove  to  be. 

a.  Human  behavior  is  neither  merely  learned  (imprinted  by  repeated 
experience),  nor  preformed  and  merely  unfolded  in  the  course  of  a 
"maturation"  process. 

b.  Human    behavior    develops    according    to    the    "ground    plan" 
(Erikson)   of  an  epigenetic  process   (of  which  libido  development  and 
ego  development  are  specific  aspects)  through  a  sequence  of  develop- 
mental crises,  whose  solution  depends  as  much  on  the  solutions   of 
previous  crises101  as  on  the  environmental  (social)  provisions  which  meet 
it  (Freud,  Hartmann,  Erikson,  Kardiner,  Sullivan). 

c.  The  laws  of  epigenesis,  whose  expression  in  the  full  perspective 
of  the  individual  life  cycle  is  the  epigenetic  "ground  plan,"  find  their 
shorter-range  expressions  in  the  regulation  of  all  behavior  and  experience 
by  intrapsychic  motivations  and  structures.  The  crucial  regulations  are 

d.  The  regulation  of  behavior  and  experience  by  motivations  and 
structures  implies :  (1)  basic  tensions  (motivations)  within  the  organism, 
which  strive  toward  reduction  and  organize  experience  and  behavior 
to  that  end;  (2)  basic  structures,  given  by  evolution,  which  on  the  one 
hand  serve  as  guarantees  of  the  organism's  adaptedness  and  adaptation 
to  the  environment  (Hartmann,  Erikson),  and  on  the  other  serve  as  the 
means  of  maintaining,  increasing,  and  discharging  the  tension  which 
exists  in  the  organism;  they  organize  experience  and  behavior  to  these 
ends;  (3)  differentiation  both  of  the  tensions  (motivations)  and  of  the 

101  Not  success  or  failure  but  the  kind  of  solution  reached  is  crucial  here  (Hart- 
mann, Erikson). 

The  Structure  of  Psychoanalytic  Theory  153 

structures,  in  such  a  manner  that  the  matrix  of  the  differentiation  sur- 
vives side  by  side  with  its  products,  though  its  manifestations  are  always 
amplified  by  these  differentiation  products;  •/4)  this  differentiation  is 
determined  both  by  the  epigenetic  laws  and  by  the  environmental  (social) 
provisions  designed  to  meet  the  epigenetic  crisis  in  which  the  differentia- 
tion in  question  comes  about;  the  differentiation  products  become 
further  guarantees  of  the  organism's  adaptedness  and  adaptation  to  the 

It  is  not  implied  that  no  other  psychoanalytic  propositions  have  this 
degree  of  generality,  nor  that  other  psychoanalytic  propositions  of  equal 
or  lesser  generality  may  not  also  have  a  long-range  survival  potential. 
These  four  propositions  summarize  that  cohesive  core  of  the  most  general 
conceptions  of  psychoanalysis  (stripped  of  their  specific  content)  which 
has  remained  constant  throughout  the  changes  of  the  theory  and  bids 
fair  to  continue  to  do  so.  It  could  be  argued  that  these  points  are  shared 
with  other  psychologies  and  are  not  specific  to  psychoanalysis.  This 
argument  does  not  hold,  though  it  is  a  clear  indication  that  psycho- 
analytic conceptions  have  been  gradually  assimilated  by  psychology  at 
large.  No  other  psychology  contains  this  assembly  of  general  conceptions, 
methods,  concepts,  and  theories;  nor  has  any  other  psychology  supported 
any  of  these  by  as  broad  an  array  of  observations  as  has  psychoanalysis. 

3.  Concepts.  The  major  concepts  of  a  high  survival  potential  per- 
taining to  each  of  the  metapsychological  points  of  view  are : 

a.  Dynamic  point  of  view.  The  concepts  of  unconscious  forces  and 
conflicts  are  close  to  observations  and  yet  of  sufficient  generality  to  have 
a  high  survival  potential.  The  concepts  of  drive,  drive-fusion,  specific 
drives  (sex,  aggression,  life  and  death  instincts,  etc.)  are  of  a  lesser  gen- 
erality, and  may  well  change  or  be  replaced  as  the  theory  changes. 

b.  Economic  point  of  view.  The  concepts  of  primary  process,  second- 
ary process,   and  pleasure  principle    (wish-fulfillment)    are  so  directly 
related  to  observation  and  so  general  that  they  are  likely  to  survive.  The 
concepts  of  cathexis,  binding,  and  neutralization,  however,  are  both  more 
inferential  and  more  specific,  and  while  the  observations  do  seem  to 
demand  some  set  of  quasi-quantitative  concepts  like  these,  it  is  uncertain 
whether  they  will  survive  in  their  present  form. 

c.  Structural  point  of  view.  The  concepts  of  structure  and  relative 
autonomy  (Hartmann)  are  indispensable  to  the  theory,  and  at  present 
it  is  not  possible  to  foresee  changes  in  the  theory  which  could  eliminate 
them.  But  the  concepts  of  id,  ego,  superego,  and  the  differentiation  of 
the  ego  into  defense-,  control-,  and  means-structures  are  neither  as  in- 
dispensable to  nor  as  independent  from  the  theory.  However,  a  variety 
of  subordinate  structural  concepts    (e.g.,  specific  primary-process  and 
defense  mechanisms,  like  displacement,  condensation,  substitution,  sym- 



bolization,  repression,  isolation,  reaction  formation,  projection)102  which 
are  more  directly  related  to  observations  and  of  a  lesser  generality,  are 
likely  to  survive.  It  is  not  implied,  however,  that  this  holds  for  all  the 
specific  defense  mechanisms. 

d.  Genetic  point  of  view.  We   discussed  above  the  high  survival 
potential  of  the  epigenetic  principle.  This  holds  also  for  the  conception  of 
the  crucial  role  of  early  experiences,  as  well  as  for  the  concepts  of  fixation 
and  regression.  The  specific  concepts  related  to  libido  development,  such 
as  orality  and  anality,  are  also  likely  to  survive  [cf.  Kardiner,  194,  195; 
Sullivan,  313;  Erikson,  56,  61,  62],  since  they  are  closely  related  to 
observations.  However,  the  classic  conception  of  libido  development  itself 
may  well  undergo  radical  change,  as  it  becomes  one  aspect  of  the  integral 
process  of  epigenesis.   The  conception  of  the  special  role  of  psycho- 
sexuality,  even  though  it  has  good  empirical  anchorage,  does  not  seem  to 
have  that  degree  of  generality  which  would  make  it  a  theoretical  neces- 
sity (cf.  pp.  89-90,  above). 

e.  Adaptive  point  of  view.  The  conceptions  of  the  organism's  pre- 
paredness for  an   average  expectable   environment    (Hartmann),    ap- 
paratuses of  primary  and  secondary  autonomy  (Hartmann),  mutuality 
(Erikson),  relative  autonomy  from  the  environment   (Gill-Rapaport), 
i.e.,  the  dependence  of  the  secondary  process  on  external  stimulation  [98, 
p,  515],  modes  and  modalities  (Erikson),  though  too  new  to  be  properly 
evaluated,  do  seem  likely  to  survive. 

Now,  in  brief,  about  the  concepts  of  the  special  (clinical)  theory.  Let 
us  take  the  transference  concept  as  an  example.  We  have  encountered 
it  as  the  foundation  of  the  long-range  significance  of  psychoanalytic 
methods.  Yet  its  own  survival  potential  might  be  characterized  as  bor- 
rowed. The  referent  of  the  transference  concept  is  not  a  single  process  but 
a  congeries  of  processes.  The  patterns  which  are  transferred  may  be  broad 
or  fragmentary,  and  the  processes  by  which  transference  is  accomplished 
are  many  and  varied:  wish-fulfillments,  displacements,  projections,  etc. 
The  transference  concept  refers  to  an  end  result:  it  is  an  achievement 
concept.  In  the  clinical  theory  of  psychoanalysis  it  is  indispensable,  but 
the  general  theory  of  psychoanalysis  resolves  it  into  process  concepts.  It 
is  probably  not  far  off  the  mark  to  suggest  that  this  is  the  case  for  most 
clinical  concepts.  A  case  in  point  is  the  very  definition  of  resistance  as 
the  manifestation  of  defense.  It  is  not  implied,  however,  that  clinical 
concepts  do  not  have  a  survival  potential :  they  do,  but  only  when  they 
are  close  to  observations  and  when  the  process-concepts  which  underlie 
them  are  themselves  likely  to  survive.  A  study  of  the  concepts  of  the 
clinical  theory  from  this  point  of  view  would  be  rewarding,  but  so  far  we 

^Objections  might  be  raised  against  discussing  these  primary-process  mech- 
anisms as  structures,  but  I  cannot  attempt  to  justify  this  here. 

The  Structure  of  Psychoanalytic  Theory  155 

do  not  have  even  a  workable  separation  of  the  special  clinical  and  the 
general  psychological  concepts  of  psychoanalysis. 


A.  Achievements 

Freud's  earliest  program  [94,  appendix]  was  to  develop  a  general 
psychology  on  neuropsychological  lines.  This  attempt  failed  and  Freud 
concluded  [94,  64]  that  the  theory  of  behavior  must  be  a  psychological 
theory.  But  he  never  gave  up  the  belief  that  once  psychoanalysis  had 
developed  far  enough,  its  link  to  physico-chemico-biological  processes 
would  be  found.  Apparently  the  time  for  this  has  not  yet  arrived  and 
the  recurrent  popularity  of  neurological  models  has  so  far  not  brought 
it  any  closer.  It  seems  that  until  psychology  has  progressed  much  further, 
attempts  at  neurological  or  biological  explanations  of  behavior  are 
bound  to  be  of  little  avail.  Freud's  program  to  develop  a  general  psy- 
chology receded  into  the  background  for  a  while,  but  was  revived  with 
the  development  of  ego  psychology. 

To  solve  the  problem  of  conversion  (somatic  compliance)  was  also 
part  of  the  earliest  program.  The  nature  of  the  hysterical  conversion 
symptom — the  psychological  conflict's  "leap  into  the  somatic" — was  and 
has  remained  a  haunting  riddle,  though  Freud  began  early  to  question 
that  psychoanalytic  methods  and  theory  could  solve  it.  Psychoanalysts, 
instead  of  solving  the  problem,  generalized  it,  first  into  the  conception  of 
"organ  neurosis"  [F.  Deutsch,  45;  Meng,  230],  and  then  into  "psycho- 
somatic medicine"  [Alexander  and  French,  4,  5,  86,  90;  Dunbar,  49,  50; 
Weiss  and  English,  321].  The  number  of  investigators  and  investigations 
in  this  field  is  great,  broad  areas  of  observation  have  been  scouted  and 
mapped,  and  the  effect  on  medicine  proper  is  considerable,  but  it  is  not 
clear  just  how  much — if  any — theoretical  advance  has  been  made.  Psy- 
chosomatic studies  remain  fraught  with  the  problem  of  "specificity," 
which  so  far  has  defied  solution.  It  is  worth  noting,  however,  that  rela- 
tively recently  a  possible  clue  to  the  conversion  riddle  has  appeared 
[see  Travell,  317]. 

The  programs  so  far  discussed  belong  to  that  phase  of  Freud's  work 
which  was  preparatory  to  psychoanalysis.  The  main  program  Freud  set 
for  psychoanalysis  proper  (1900)  was  to  explore  the  unconscious;  later 
(1923)  this  changed  into  the  exploration  of  the  id  and  the  unconscious 
ego.  Discoveries  are  still  being  made  in  both  areas  and  much  of  the  "un- 
conscious ego"  is  still  uncharted  territory.  Yet  considering  that  successful 
exploration  always  breeds  new  problems,  the  work  on  this  program 


can  be  considered  well  advanced.  This,  however,  is  a  judgment  within 
the  frame  of  reference  of  psychoanalysis:  it  refers  to  the  program  of  ex- 
ploration and  not  to  a  program  of  testing  and  developing  the  theory  by 
means  of  experiments. 

A  related  program  was  to  apply  the  theory  to  myth,  legend,  literature, 
art,  ethnology,  etc.,  in  order  to  demonstrate  its  pertinence  to  all  human 
behavior  and  products  and  thus  to  obtain  a  broad  base  of  supporting 
evidence.  The  achievements  in  this  direction  have  already  been  men- 
tioned, and  the  importance  of  the  new  means  provided  by  ego  psy- 
chology for  the  further  pursuit  of  this  program  has  been  indicated. 

For  a  long  while  the  exploration  of  the  ego  seemed  to  be  only  a 
contemplated  program.  Freud  expected  the  information  about  ego 
functions  to  come  from  the  study  of  "narcissistic  neuroses"  (i.e.,  psy- 
choses), but  delayed  this  study  because  he  considered  the  exploration  of 
the  unconscious  to  be  the  primary7  task.  Yet  this  program  was  indirectly- 
pursued  throughout  the  history  of  psychoanalysis  in  the  study  of  the 
defenses,  censorship,  secondary  process,  and  reality  relationships.  How- 
ever, Freud  did  embark  (1921,  1923)  on  an  explicit  conceptualization 
of  the  ego  without  studying  psychoses  anew,  apparently  prompted  by  the 
problem  of  "the  negative  therapeutic  reaction53  and  "the  unconscious 
sense  of  guilt"  [124,  126].  Later  he  carried  the  study  of  the  ego  further 
(1926)  by  re-evaluating  the  problem  of  anxiety  [131].  Other  psycho- 
analysts followed  his  lead  [77,  245,  284,  320]  and  the  achievements  of 
this  phase  of  the  program  were  capped  (1936)  by  A.  Freud's  [93]  work. 
The  ego-psychological  program  was  then  dramatically  broadened  by 
Hartmann  [157]  and  by  Erikson  [56,  59].  The  ego  was  explored  slowly 
but  so  successfully  that  a  broad  and  still  uncharted  area  was  opened  up. 

The  program  of  superego  exploration  was  already  implicit  in  the 
study  of  censorship  (1900).  But  only  the  study  of  narcissism  [114] 
brought  it  into  focus  in  the  term  "ego  ideal"  (1914).  Though  the  con- 
cept of  the  superego  was  formalized  simultaneously  with  the  ego  [124, 
126],  and  in  spite  of  significant  advances  [see  73],  the  work  on  this  pro- 
gram has  hardly  passed  the  beginning  stages. 

While  psychoanalysis  as  a  therapy  is  primarily  the  subject  matter 
of  the  special  (clinical)  theory,  the  theory  of  therapeutic  technique  is 
part  of  the  general  theoretical  program  of  psychoanalysis.  It  was  so 
treated  by  Freud  in  the  prehistory  of  psychoanalysis  [35,  chap.  3],  in 
some  of  the  "Papers  on  Technique"  [106]  and  in  "Analysis  Terminable 
and  Interminable"  [139].  Nevertheless,  this  program  is  still  far  from  ful- 
fillment. Even  the  most  systematic  [72]  of  the  few  extensive  [151,  224] 
treatments  of  technique  contributes  little  toward  the  theoretical  pro- 
gram. E.  Bibring  [28]  has  penetrated  into  these  problems  further  than 
most  others.  Recently  Eissler  [51]  and  Gill  [145]  have  also  made  rele- 

The  Structure  of  Psychoanalytic  Theory  157 

vant  contributions.  The  importance  of  this  program  and  the  difficulties 
in  its  way  are  equally  great.  Progress  may  require  not  only  the  study  of 
the  techniques  of  psychoanalysis  and  those  of  other  schools  of  therapy, 
but  also  the  development  of  a  psychoanalytic  theory  of  communication. 

The  theoretical  explanation  of  neuroses  was  an  outstanding  part  of 
the  program.  This  is  where  the  work  on  the  special  (clinical)  theory  of 
psychoanalysis  had  its  greatest  achievements  and  also  brought  a  con- 
siderable general  theoretical  yield,  to  which  FenichePs  [735  77]  systematic 
survey  of  the  special  theory  refers  continuously.  Yet  we  still  do  not  have 
a  systematic  treatment  of  neuroses,  from  the  point  of  view  of  the  general 
theory,  comparable  to  that  which  Freud  gave  of  dreams  in  chapter  7 
of  The  Interpretation  of  Dreams  [98]. 

The  theoretical  explanation  of  psychoses  was  also  a  part  of  the  pro- 
gram. Beginning  with  his  early  (1895)  study  of  a  case  of  paranoia  [941 
and  his  analysis  of  the  Schreber  case  [107],  Freud  dwelt  on  it  repeatedly 
[114,  117,  119,  126,  127,  128].  Yet  despite  the  contributions  of  Abraham 
[1],  Federn  [70],  B.  Lewin  [2141,  Fromm-Reichmann  [142,  143], 
M.  Wexler  [324,  325],  Hartmann  [163],  and  others,  and  in  spite  of  the 
studies  by  Putnam,  Mahler,  Bettelheim,  and  other  psychoanalysts  on 
juvenile  schizophrenia,  the  fulfillment  of  this  program  has  barely  begun. 

The  situation  is  only  slightly  better  in  that  part  of  the  program  which 
comprises  the  general  theory7  of  character  disorders,  addictions,  delin- 
quency, criminality,  and  borderline  problems  [22,  23,  202,  203,  282, 
283].  J 

B.  Convergence  with  Other  Theories 

It  is  difficult  to  differentiate  the  applications  of  the  theory  to  other 
fields,  its  influence  on  other  sciences,  and  its  convergence  with  other 
theories  and  sciences.  The  distinction  might  be  drawn,  perhaps,  as 
follows:  application  is  the  work  of  psychoanalysts  in  other  fields;  in- 
fluence is  the  adoption  of  psychoanalytic  assumptions,  methods,  findings, 
and/or  theories  by  workers  in  other  fields;  convergence  is  mutual  in- 

In  this  sense,  in  anthropology  the  days  of  application  [Freud,  109; 
Roheim,  289,  291,  etc.]  and  influence  (e.g.,  Kluckhohn)  are  past,  and 
convergence  can  be  observed  on  one  side  in  Erikson's  work,  and  on  the 
other  in  the  work  of  cultural  anthropologists  (however  the  opinions  may 
vary  about  this  work  otherwise).  [See  also  Psychoanalysis  and  the  Social 
Sciences,  290.] 

The  same  holds  for  sociology,  where  the  days  of  application  and  in- 
fluence (Freud,  W.  Reich,  Fenichel,  Lasswell,  and  the  early  Fromm) 
are  past  and  convergence  can  be  observed,  for  instance,  in  Parsons', 
Riesman's,  and  N.  Foote's  work  on  the  one  hand  (however  the  opinions 


may  vary  about  this  work  otherwise),  and  in  Hartmann's  and  Erikson's 
on  the  other. 

The  convergence  of  psychoanalysis  with  medicine  in  general  and 
psychiatry  in  particular,  though  only  too  obvious,  is  practical  rather  than 

It  is  questionable  whether  one  can  speak  of  a  convergence  of  psycho- 
analysis with  the  other  fields,  its  applications  to  which  wrere  mentioned 
earlier:  in  art,  literature,  history,  etc.,  we  find  influence,  but  no  more. 

Now,  to  the  convergence  of  psychoanalysis  with  psychology.  We  have 
already  mentioned  that  psychoanalytic  ego  psychology  seems  to  be  re- 
sponsible for  a  considerable  part  of  this.  The  convergence  with  develop- 
mental psychology  is  of  long  standing  (Werner  and  the  early  Piaget) 
and  is  reinforced  on  the  one  hand  by  the  recent  longitudinal  and  cross- 
sectional  studies  of  psychoanalysts  like  M.  Fries,  Spitz,  Escalona  and 
Leitch,  Benjamin,  Kris,  Mittelmann;  and  on  the  other  by  the  studies  of 
Piaget,  Werner  and  his  associates,  and  others.  The  work  of  K.  Lewin  and 
his  associates  on  the  one  hand,  and  that  of  T.  French  on  the  other  are 
outstanding  indications  of  convergence.  The  studies  in  learning  theory 
by  Bollard,  Miller,  Mowrer,  Sears,  etc.  (however  the  opinions  may  vary 
about  this  work  otherwise),  represent  a  convergence  of  psychology  with 
psychoanalysis.  Murray's  early  work,  his  and  his  associates'  and  their 
successors'  work  in  the  assessment  of  personality  are  also  indications  of 
this  convergence;  so  is  much  of  the  recent  work  in  experimental  clinical 
psychology.  Two  other  important  indications  of  convergence,  the  studies 
on  motivations  and  memory  and  on  motivations  and  perception,  have 
already  been  discussed. 

The  future  of  this  convergence  may  hinge  on  the  solution  of  the 
problem  of  structure  formation  and  learning.  If  that  solution  should 
arise  from  the  matrix  of  psychoanalytic  theory,  the  latter  may  become 
the  core  of  psychology  proper.  If  the  solution  should  prove  relatively 
independent  of  psychoanalysis,  then  the  latter  is  likely  to  become  a  rela- 
tively subordinate  part  of  the  general  theory  of  psychology  as  the  core 
of  its  clinical  and  motivational  theories,  but  its  concepts  and  theories  will 
be  reducible  to  more  fundamental  ones.  The  existing  learning  theories 
have  not  accomplished  this  reduction  and  it  seems  unlikely  that  they  ever 

Finally,  coming  closest  to  the  home  base  of  psychoanalysis,  the  devel- 
opment of  psychoanalytic  ego  psychology  has  begun  to  extract  the  valid 
contributions  from  the  theories  of  the  Neo-Freudian  schools,  and  thus 
to  initiate  the  convergence  of  these  offshoots  with  psychoanalytic  theory 
proper.  There  is  still  much  to  be  done  here  and  the  convergence  pertains 
only  to  the  theories,  not  to  the  "schools"  as  organizations  of  vested 

The  Structure  of  Psychoanalytic  Theory  159 



A.  Empirical  Evidence  Needed 

Psychoanalytic  theory  does  not  need  additional  data  per  se  for  its  de- 
velopment: the  amount  of  data  is  already  embarrassingly  large.  It  does 
not  need  the  Blacky  Test  type  of  data  which,  though  amenable  to 
statistical  treatment,  are  simply  masked  clinical  data:  the  clinical  data 
are  better.  It  does  not  need  experimental  data  which  replicate  clinical 
relationships.  What  it  needs  are  methods  to  obtain  data  which  can  lead 
beyond  the  clinical  relationships  to  theoretical  relationships  of  the  type 
discussed  in  this  essay. 

It  is  for  this  reason  that  experimental  data  on  structure  formation 
and  learning,  and  data  corroborating  or  negating  Piaget's  observations 
and  theories  are  needed. 

In  Hartmann's  theory,  we  have  for  the  first  time  a  concept  of  auton- 
omy and  we  have  (cf.  pp.  95ff.  and  p.  98,  above)  also  an  elaboration 
of  it  into  a  threefold  conception  of  relative  autonomy  (Erikson,  Gill). 
The  methods  used  in  the  McGill  University  and  Bethesda  studies  on 
sensory  deprivation,  as  well  as  the  hypnotic  methods,  seem  to  be  ap- 
propriate means  to  alter  the  balance  of  these  relative  autonomies  [see 
280].  The  theory  needs  data  obtained  by  these  or  other  relevant  methods 
(e.g.,  drugs  like  mescaline),  but  just  any  data  obtainable  by  these 
methods  will  not  do:  the  need  is  for  data  obtained  in  controlled  experi- 
ments guided  by  the  theory  of  autonomy. 

In  Hartmann's  studies  we  have  for  the  first  time  a  concept  of  conflict- 
free  ego  functions.  G.  S.  Klein's  work  contains  a  variety  of  methods  for 
the  study  of  those  ego  structures  which,  unlike  defenses,  are  conflict  free 
and  serve  to  control  and  channel  motivations.  Data  concerning  such 
structures  are  needed.  But  an  indiscriminate  proliferation  of  such  data 
would  provide  mainly  a  catalogue  of  "cognitive  attitudes"  (Klein),  just 
as  French  psychiatry  has  provided  us  with  a  term  for  more  or  less  every 
possible  form  of  phobia.  The  data  needed  are  those  which  will  elucidate 
the  relation  of  these  style  structures  to  other  ego  structures  (e.g.,  de- 
fenses), to  motivations,  and  to  each  other. 

In  Erikson's  work  we  have  for  the  first  time  a  theory  and  an  epi- 
genetic  ground  plan  of  ego  development.  Additional  data  concerning 
each  epigenetic  phase  are  needed.  But  again,  just  any  data  pertaining  to 
a  given  phase  of  development  will  not  do.  The  data  should  pertain  to 
Erikson's  observations  and  should  corroborate,  elaborate,  modify,  or 
negate  them.  To  obtain  such  data,  the  investigators  will  have  to  adopt 
Erikson's  frame  of  reference,  at  least  to  begin  with. 

We  suggested  above  that  a  prerequisite  of  the  theory  of  therapeutic 


technique  may  be  a  theory  of  communication.  Data  for  building  such  a 
theory  are  needed.  The  data  and  concepts  of  the  existing  attempts  at  a 
communication  theory  do  not  seem  to  be  relevant.  The  focus  of  such  a 
communication  theory7  must  be  the  laws  which  govern  the  tendency  of 
communication  to  engender  or  to  prevent  reciprocal  communication. 
Moreover,  it  should  be  a  theory7  in  which  the  communicants'  becoming 
conscious  of  something  is  equivalent  to  (latent)  verbal  or  nonverbal 
communication  [see  263,  273].  The  methods  by  which  data  relevant 
to  such  a  theory  can  be  obtained  have  yet  to  be  worked  out. 

In  Hartmann's  "self  [161,  162]'  and  in  Erikson's  "identity"  [66] 
we  have  in  psychoanalytic  theory  for  the  first  time  concepts  to  account 
for  the  historical  continuity  of  the  individual  and  for  his  self-experience, 
and  conceptual  tools  to  distinguish  them  from  the  referents  of  the  ego 
concept.  Data  pertaining  to  and  permitting  the  elaboration  of  these  con- 
cepts are  needed.  But  just  any  data  of  "self-experience,"  "self-evalua- 
tion," or  "ego-involvement"  will  not  do.  They  must  be  data  concerning 
the  relation  of  the  "self5  or  of  "identity"  to  the  psychoanalytic  theory  of 
psychological  functions  in  general  and  of  ego  functions  in  particular. 

The  less  than  satisfactory  progress  in  the  theoretical  understanding 
of  schizophrenia  and  other  psychoses  has  been  mentioned.  Here  again 
data  are  in  abundance.  What  to  do  with  them  is  the  question.  They  have 
not  been  selected  to  reveal  the  relation  of  the  phenomena  of  schizo- 
phrenia to  the  existing  theory.  There  is  no  need  for  more  data  showing 
that  the  content  of  psychotic  products  can  be  interpreted  like  dreams  or 
unconscious  fantasies.  Nor  are  data  needed  on  oral  or  anal  wishes  under- 
lying the  manifest  content  of  psychotic  products;  these  are  ubiquitous 
in  man,  and  only  their  role,  intensity,  and  frequency  might  conceivably 
be  specific  to  a  given  psychosis.  It  is  the  formal  characteristics  of  psy- 
chotic behavior  (action,  affect,  and  thought)  which  seem  to  be  specific, 
and  what  is  needed  are  data  to  connect  them  with  the  psychoanalytic 

Data  are  needed  to  reveal  the  similarities  and  differences  between 
analogous  structures  (and  motivations)  on  different  hierarchic  levels  of 
the  psychological  organization.  J.  F.  Brown  [36]  obtained  some  data  of 
this  kind  [cf.  also  276]. 

Last  but  not  least,  though  no  data  replicating  clinical  relationships 
are  needed,  any  replication  whose  purpose  is  to  quantify  these  relation- 
ships so  as  to  pave  the  way  toward  "dimensional  quantification"  should 
be  welcome. 

This  enumeration  has  no  systematic  pretensions,  nor  does  its  sequence 
imply  an  order  of  importance.  The  examples  were  chosen  to  show  that, 
more  than  data,  we  need  methods  which  promise  to  yield  data  relevant 
to  the  theory  and  its  unsolved  problems. 

The  Structure  of  Psychoanalytic  Theory  161 

B.  Obstacles  to  the  Development  of  the  Theory 

Here  we  will  dwell  on  theoretical  obstacles,  leaving  the  discussion  of 
practical  obstacles  for  Section  C. 

The  days  of  the  resigned  belief  that  complex  psychological  phenom- 
ena cannot  be  studied  in  the  laboratory  are  past.  So  is  the  overenthu- 
siasm  of  K.  Lewin's  early  days  [220],  when  there  seemed  to  be  no  doubt 
that  all  psychological  phenomena  could  be  relevantly  studied  on  the 
laboratory  scale.  While  we  all  hope  that  even-  psychological  phenom- 
enon Is  amenable  to  scientific  study,  to  find  the  ways  and  means  for  this 
has  become  our  gravest  concern.  The  mam  obstacles  to  the  development 
of  psychoanalytic  theory  center  around  these  ways  and  means. 

First  y  due  regard  for  the  individual's  rights  sets  limits  to  the  manipu- 
lation of  beha\ior  outside  and  even  inside  the  laboratory;  and  due  regard 
for  the  privacy  of  the  individual  sets  limits  even  to  observation.  This  Is 
one  of  the  major  empirical  obstacles.  The  problem  Is  not  only  the  ethical 
one  of  trespassing  on  rights  and  privacy  but  also,  and  perhaps  primarily, 
what  such  trespassing  does  to  the  subject,  to  the  observer,  and  to  the 

Second,  the  hierarchic  problem,  so  heavily  stressed  In  these  pages, 
implies  that  reduction  to  laboratory  size  more  often  than  not  changes 
the  hierarchic  position  of  the  phenomenon  or  relationship  in  question, 
so  that  not  the  phenomenon  or  relationship  itself,  but  a  high-level  hier- 
archic equivalent  of  it  is  studied.  This  is  not  simply  an  obstacle.  It  indi- 
cates that  laboratory  research  can  attack  all  psychological  problems, 
provided  it  centers  its  attention  on  the  laws  of  hierarchic  transformations. 
Once  such  laws  begin  to  take  shape,  psychologists  will  be  able  to  dis- 
pense with  the  arbitrary  claim  that  the  laboratory  findings  obtain  for 
life  situations  and  will  use  these  laws  as  the  rules  by  which  inferences 
from  the  laboratory  findings  to  life  situations  can  be  drawn.  This  the- 
oretical complexity  then  is  not  per  se  an  obstacle,  though  there  is  long 
and  arduous  experimentation  ahead  before  these  laws  of  hierarchic  rela- 
tions are  discovered  and  brought  to  a  point  where  they  can  serve  as  rules 
of  inference. 

Third,  laboratory  methods  cannot  get  around  the  troublesome  fact 
that  there  are  many  psychological  phenomena  which  occur,  as  a  rule, 
only  in  the  contact  of  one  person  with  another  (or  others).  The  study  of 
such  phenomena  led  to  the  method  of  participant  observation  in  therapy, 
in  everyday  life,  and  in  laboratory  situations.  This  method  has  scarcely 
been  explored  theoretically;  in  it  the  investigator  enters  into  the  privacy 
of  the  subject,  but  he  does  so  at  the  price  of  becoming  a  participant, 
shouldering  all  those  implicit  and  explicit  commitments  which  participa- 
tion involves.  Psychoanalysts  and  other  therapists  know  a  great  deal 


about  these  commitments  and  their  effect  on  the  observer  and  on  the  ob- 
served. But  the  implications  of  this  knowledge  for  the  method  have  not 
yet  been  theoretically  formulated  [see,  however,  Bemfeld,  14,  Gross, 
153],  and  the  lack  of  such  systematization  is  an  obstacle  in  the  way  of 
the  theory's  development. 

We  have  discussed  the  possibility  that  experimental  study  will  dis- 
cover rules  of  inference,  by  means  of  which  conclusions  can  be  drawn 
from  laboratory-sized  to  life-sized  phenomena.  What  about  the  rules  of 
inference  for  relating  data  obtained  by  direct  observation  to  data  ob- 
tained from  participant  observation?  For  instance,  the  psychoanalytic 
theory  of  development  is  built  from  reconstructions  based  on  data  ob- 
tained by  the  method  of  interpersonal  participant  observation  in  the 
therapeutic  two-group  situation,  while  Piaget's  theory  of  development 
is  built  on  data  obtained  in  direct  observation.103  Now  it  is  possible  that 
the  theories  of  development  of  psychoanalysis  and  of  Piaget  will  prove 
compatible,  and  rules  of  inference  will  be  found  to  link  their  concepts. 
Indeed,  it  is  possible  that  their  mutual  influence  will  lead  to  a  redefini- 
tion of  their  concepts  so  that  there  will  arise  a  single  conceptual  system 
which  subsumes  both  theories  or  subsumes  one  under  the  other.  But 
there  are  two  other  possibilities.  First,  the  two  theories  might  prove  in- 
compatible and  thus  one  of  them  untenable.  Second,  it  might  just  hap- 
pen that  the  two  theories,  like  the  observations  they  are  based  on,  will 
prove  not  to  overlap,  and  not  to  be  incompatible.  If  so,  the  two  methods 
will  have  arrived  at  theories  pertaining  to  two  different  aspects  of  the 
same  subject  matter.  We  might,  then,  have  to  conclude  that  these  two 
aspects  of  the  subject  matter  are  complementary  [cf.  Niels  Bohr's  com- 
plementarity concept  in  atomic  physics,  and  the  complementarity  in  the 
study  of  the  living  cell  envisaged  by  the  biophysicist  Delbrueck,  44]. 
The  uncertainty  whether  the  yield  of  the  participant  observation  method 
and  the  yield  of  other  methods  can  be  related  to  each  other  by  conjunc- 
tive rules  of  inference,  or  must  be  related  by  a  disjunctive  rule  of  comple- 
mentarity, is  a  major  hurdle  in  the  way  of  the  development  of  psycho- 

10SA11  observations  on  human  beings  are  in  a  sense  participant  observations: 
one-way  screens,  movies,  and  sound  tracks  obscure  but  do  not  circumvent  this  fact. 
Yet  there  is  a  difference  between  being  a  participant  observer  and  using  the 
method  of  interpersonal  participant  observation,  and  there  is  also  a  difference 
between  constructing  and  reconstructing  developmental  relationships  from  partici- 
pant observations.  Piaget  [254,  255,  256],  too,  was  a  participant  observer:  by  his 
actions,  he  modified  the  situations  and  the  tasks  his  children  faced.  Yet  he  did 
not  use  the  method  of  participant  observation,  in  that  he  did  not  systematically 
study  the  changes  in  the  children's  relation  to  him  (their  father)  consequent  to 
his  "participation/'  nor  the  changes  in  their  sensorimotor  behavior  as  the  latter 
depended  on  the  children's  relation  to  him. 

The  Structure  of  Psychoanalytic  Theory  163 

analytic  theory.  It  is  possible  that  an  inkling  of  this  difficulty  accounted 
for  Freud's  lack  of  interest  in  the  attempts  to  verify  psychoanalytic  prop- 
ositions by  methods  other  than  psychoanalytic. 

Fourth,  the  last  of  the  obstacles  to  be  mentioned  here  is  the  problem 
of  mathematization,  including  quantification,  already  discussed.  It  is 
both  an  empirical  and  a  theoretical  obstacle  to  the  development  of  the 

C.  The  Practical  Obstacles  to  Theoretical  Advance  in  Psychology 

Let  us  first  take  the  obstacles  in  the  way  of  psychoanalytic  theory. 
Here  the  lack  of  systematic  theoretical  literature,  the  nature  of  psy- 
choanalytic training,  and  the  character  of  psychoanalytic  practice  stand 

As  a  rule,  the  observer  and  experimenter  is  guided  in  his  contribu- 
tion to  theory  development  by  systematic  theoretical  literature.  In  the 
lack  of  such,  the  investigator  has  to  master  the  primary  sources  and  do 
the  systematization  for  himself.  This  is  a  time-consuming  pursuit  to 
which  the  habits  of  psychoanalytic  practice  are  not  conducive.  It  is  often 
said  that  psychoanalytic  theory  is  a  rigid  and  unchangeable  doctrine. 
Although  there  is  such  dogmatism  and  orthodoxy  in  the  Societies  and 
Institutes  (whether  they  are  Freudian  or  Neo-Freudian  or  in  between) 
in  regard  to  the  clinical  theory,  I  have  rarely  found  dogmatism  in  regard 
to  the  general  theory.  The  attitudes  range  from  enthusiasm,  through  lack 
of  interest,  to  total  lack  of  information.  The  general  theory,  far  from 
being  well-ingrained  dogma,  is  a  waif  unknown  to  many,  noticed  by 
some,  and  closely  familiar  to  few.  Not  the  alleged  rigidity  of  the  theory, 
but  rather  unfamiliarity  with  it  is  the  obstacle  to  theoretical  progress. 
The  lack  of  systematic  theoretical  literature  is  certainly  not  the  sole  cause 
of  this  situation  (the  original  sources  are  available)  but  it  is  a  major 
handicap  to  advancement. 

The  training  given  by  psychoanalytic  Institutes  is  primarily  designed 
for  future  practitioners,  and  limited  to  physicians.  The  scope  of  this 
training  is  defined  by  several  factors:  (a)  its  "night  school"  character, 
(b)  the  average  medical  training,  which  prepares  the  students  neither 
for  psychology,  psychiatry,  and  psychoanalysis,  nor  for  theoretical  and 
research  pursuits,  (c)  the  fact  that  both  teachers  and  students  are,  as  a 
rule,  full-time  practitioners.  Two  additional  facts  about  this  training: 
first,  it  is  postdoctoral,  time-consuming,  and  costly,  and  thus  pushes  the 
graduate  to  seek  more  lucrative  and  less  leisurely  pursuits  than  research; 
second,  though  the  rules  limit  it  to  physicians,  some  psychologists  and 
other  scientists  can  obtain  "research  training"  in  psychoanalysis,  but  this 
includes  only  training  analysis  and  course  work,  not  supervision  (control) 



and  often  not  even  clinical  seminars,104  although  the  theoretician  and 
research  man  needs  full  training  no  less  than  does  the  future  practitioner. 
Thus  the  "medical  closed  shop55  works  doubly  against  progress  in  psycho- 
analytic theory.  It  is  small  wonder  that  divergences  of  observation  and 
thinking  among  psychoanalysts  tend  to  be  resolved  not  by  theoretical 
or  empirical  decision  but  by  orthodoxy  and  secessions. 

The  nature  of  psychoanalytic  practice  does  not  foster  theoretical 
development.  The  long  workdays,  spent  closeted  with  patients,  provide 
neither  the  necessary  time  and  leisure  nor  the  detachment.  The  solitary 
character  of  the  practice  minimizes  that  kind  of  collegial  interchange 
which  is  the  fertile  soil  of  theory  making.  The  grants  available  and  the 
institutions  which  relieve  some  psychoanalysts  from  the  burden  of  full- 
time  practice  and  provide  opportunities  for  such  interchange  are  for  the 
privileged  few.  While  no  science  has  more  than  a  few  theoreticians  at 
a  time,  those  few  always  emerge  from  the  many  who  try.  Where  only  a 
few  can  try,  the  prospects  remain  dim,  however  well  or  poorly  the  few 
may  be  chosen. 

The  effect  on  the  psychoanalyst  of  the  limitation  on  the  number  of 
patients  he  can  see  is  enhanced  by  the  limited  range  of  people  con- 
sidered treatable  by  psychoanalysis  and  by  the  limited  number  who  can 
afford  it.  Moreover,  the  outstanding  psychoanalysts  sooner  or  later  be- 
come training  analysts,  and  then  part  of  their  time  is  occupied  with  an 
even  more  limited  group:  the  kind  of  people  who  want  to  become  psy- 
choanalysts and  pass  through  the  sieve  of  the  training  committees.  These 
limitations  are  particularly  crippling  to  the  development  of  the  psycho- 
social  aspects  of  the  theory,  but  they  also  leave  psychoanalysis  centered 
on  its  clinical  aspects,  to  the  neglect  of  its  general  theory.  True,  the  clin- 
ical theory  needs  further  development  and  its  methods  are  so  far  indis- 
pensable for  the  study  of  a  wide  range  of  phenomena;  but  there  is 
another  wide  range  of  phenomena,  crucial  for  the  development  of  the 
general  theory,  that  is  not  amenable  to  study  in  the  therapeutic  situation 
where  the  patient's  interest  is — and  should  be — the  guide.  The  develop- 
ment of  ego  psychology  is  particularly  affected  by  this  limitation. 

It  seems  that  without  more  scholarly  and  academic  training,  and 
without  the  admission  of  nonmedical  students  to  such  training,  the  main 
obstacles  in  the  way  of  the  development  of  psychoanalytic  theory  are 
bound  to  persist.  It  is  unlikely  that  medical  schools  or  psychology  de- 
partments would  do  better  than  the  psychoanalytic  Institutes:  neither 
their  traditions,  nor  their  chances  of  recruiting  training  staffs,  nor  the 
complexity  of  the  training  problem  to  be  met  seems  to  bode  well  for 
such  "simple"  solutions. 

104  Since  this  was  written,  initial  steps  have  been  taken  by  the  American 
Psychoanalytic  Association  to  explore  ways  to  change  this  situation. 

The  Structure  of  Psychoanalytic  Theory  165 

Now,  the  obstacles  to  theoretical  progress  in  psychology:  the  "sci- 
entific method,"  the  addiction  to  a  single  method  (or  limited  set  of 
methods) ,  and  the  measuring  rage  stand  out. 

Theory  making,  i.e.,  theoretical  progress,  begins  in  familiarity  with 
phenomena  and  in  thinking  about  them  for  about  the  theories  pertain- 
ing to  them).  It  continues  in  hunches  and  speculations,  some  of  which 
are  amenable  to  empirical  test;  others,  which  spin  relations  between  con- 
cepts and  theories,  or  restructure  and  systematize  them,  are  not  and  need 
not  be,  though  they  may  well  lead  to  conclusions  \vhkh  again  can  and 
must  be  subjected  to  empirical  test. 

The  "scientific  method"  is  the  canon  by  which  that  record  is  made 
which  we  call  science  the  codified,  interconnected  body  of  accepted 
knowledge.  But  it  is  not  the  canon  for  making  discoveries,  nor  the  canon 
for  making  theories.  Nor  is  the  canon,  by  which  the  scientific  record  is 
made,  unique  and  static:  it  changes  with  the  change  in  the  methods, 
subject  matter,  and  aims  of  research.  Dingle,  the  British  historian  of 
science,  had  harsher  words  about  the  "scientific  method,  or  methodol- 
ogy as  it  is  often  called  now" : 

...  a  discipline  conducted  for  the  most  part  by  logicians  unacquainted 
with  the  practice  of  science,  and  it  consists  mainly  of  a  set  of  principles  by 
which  accepted  conclusions  can  best  be  reached  by  those  who  already  know 
them.  When  we  compare  these  principles  with  the  steps  by  which  the  dis- 
coveries were  actually  made  we  find  scarcely  a  single  instance  in  which  there 
is  the  slightest  resemblance.  If  experience  is  to  be  any  guide  to  us  at  all — and 
what  scientist  can  think  otherwise — we  must  conclude  that  there  is  only  one 
scientific  method:  produce  a  genius  and  let  him  do  what  he  likes  .  .  .  the 
best  we  can  do  is  to  learn  to  spot  natural  genius  .  .  .  and  protect  it,  by 
fiery  dragons  if  need  be,  from  the  god  of  planning  [46,  pp.  38-39]. 

Beveridge  [24]  described  scientific  investigation  as  an  art.  Theory 
making  may  be  described  as  a  work  of  imagination;  the  "scientific 
method"  comes  into  play  only  in  testing  the  theory  and  in  making  the 
record.  But  even  there,  however  much  the  scientific  method  can  help  to 
design  economic  and  valid  tests,  the  essential  ingredient  is  still  the  in- 
genuity in  inventing  a  method  which  connects  the  phenomena  and  the 

The  stress  on  the  "scientific  method"  becomes  an  obstacle  to  the- 
oretical advance  in  several  ways.  First,  the  stress  on  teaching  the  scien- 
tific method  and  the  design  of  experiment  diverts  attention  from  training 
in  observation.  Second,  it  discourages  the  budding  investigator's  interest 
and  trust  in  his  own  hunches  and  speculations.  Third,  it  makes  the  "sci- 
entific method"  and  the  "design  of  experiment"  appear  as  a  sure-fire 
way  to  produce  "research  findings."  The  findings  thus  produced  clutter 
our  literature  and  crowd  out  the  interest  in  methods  of  experimenting 


and  observing.  Fourth,  it  leads  to  a  publication  policy  (and,  through  it, 
to  a  training  by  precept)  such  that  the  publications  conform  to  the  "sci- 
entific method55  and  cover  up  the  actual  tracks  of  the  investigator  even 
when  by  chance  his  tracks  would  be  worth  knowing.  The  publications 
read  as  though  investigation  consists  of  nothing  but  the  application  of 
the  scientific  method.  Thus  to  the  novice,  our  (and  what  is  more  im- 
portant, his  own)  actual  disorderly  ways  of  productive  thinking  appear 
as  an  inadequacy.  His  self-observations,  which  show  him  that  his  think- 
ing does  not  follow  the  "scientific  method,"  become  the  sources  of  a 
gnawing  self-doubt,  which  in  turn  only  too  often  leads  to  a  sterilizing 
discipline  of  thought.  No  wonder  that  in  our  literature  few  authors  are 
surprised,  few  things  are  surprising,  and  a  deadly  boredom  prevails,  aided 
and  abetted  by  what  the  given  journal  considers  to  be  the  form  of  sci- 
entific reporting. 

The  bane  of  the  "single  theory  and  single  method"  is  in  part  synon- 
ymous with  the  plague  called  "schools  of  psychology."  The  investigator 
uses  a  method  and  becomes  its  captive.  So  do  his  students.  He  develops 
a  theory  which  can  only  predict  phenomena  elicited  by  that  method  or 
a  closely  related  one.  What  is  not  amenable  to  study  by  those  methods 
ceases  to  influence  the  theory.  In  turn,  all  theories  whose  methods  do  not 
apply  to  the  realm  of  phenomena  in  question  are  somehow  considered 
"wrong,"  and  if  they  are  tested  at  all,  it  is  by  methods  alien  to  them, 
and  so  they  are  obviously  found  wrong.  Usually,  however,  they  are  ig- 
nored altogether.  As  a  result,  certain  methods  become  "canonized,"  the 
study  of  a  limited  range  of  phenomena  becomes  the  only  "proper  study 
of  man,"  and  those  who  try  to  reunite  the  field  of  psychology,  so  frag- 
mented by  a  few  methods,  are  regarded  as  "philosophers"  in  the  pejora- 
tive sense  of  the  word.  To  be  a  theorist  becomes  an  opprobrium :  this  is 
the  particular  form  of  anti-intellectualism  which  is  endemic  in  present- 
day  psychology.  No  new  methods  (i.e.,  ways  of  experimenting,  in  contra- 
distinction to  designs  of  experiment)  are  sought  to  break  the  splendid 
isolation  of  the  self-encapsulated  realms  of  phenomena  thus  created. 
Methodological  thinking,  which  deals  with  the  relation  of  method  and 
theory,  and  attempts  to  establish  what  is  an  artifact  of  the  investigative 
method  and  what  is  "the  nature  of  the  beast,"  remains  mostly  beyond 
the  ken  of  the  psychologist.105 

The  "measuring  rage,"  already  discussed,  is  particularly  character- 
istic of  the  experimental  work  in  clinical  and  personality  psychology.  It 
expresses  and  fosters  a  disregard  for  theory,  and  is  thus  a  major  obstacle 

300  This  methodological  implication  of  Brunswik's  "representative  sampling  of 
design"  is  often  overlooked.  For  brevity  and  emphasis,  I  deliberately  overstate 
these  points:  for  instance,  the  "artifact"  issue  is  by  no  means  as  simple  as  the 
above  statement  suggests. 

The  Structure  of  Psychoanalytic  Theory  167 

to  theoretical  advancement.  But  It  also  distracts  attention  from  the  gen- 
eral problem  of  mathernatlzation  and  the  specific  problem  of  dimensional 
quantification.  We  may  not  be  too  far  off  the  mark  In  suggesting  that  the 
malaise  of  psychology  which  is  manifested  In  the  "measuring  rage"  is  the 
same  as  the  one  responsible  for  the  epidemic-like  popularity  in  psychol- 
ogy of  ''information  theory,"  "open  systems/5  "stress  syndrome,"  and 
other  extrapsychological  achievements.  Conceptions  and  methods  can 
be  borrowed  from  other  sciences:  all  that  Is  useful  should  be  used.  But 
the  epidemic  of  grasping  at  every  likely  new  achievement  of  other  sci- 
ences seems  to  be  a  symptomatic  giveaway:  salvation  is  expected  from 
the  outside  and  not  from  results  achieved  by  the  sweat  of  our  own 
brows.  At  the  root  of  It  Is  a  lack  of  self-confidence :  the  lack  of  assurance 
that  psychology  knows  where  It  has  come  from  and  where  It  is  going. 


1.  Abraham,  K.  Selected  papers  of  .  .  .  .  London:  Hogarth.  1927. 

2.  Adams,  D.  K.  A  restatement  of  the  problem  of  learning.  Brit.  J. 
Psychol.,  1931,  22,  150-178. 

3.  Alexander,   F.   The  relation  of  structural  and  instinctual  conflicts. 
Psychoanal  Quart. ,1933,  2,  181-207. 

4.  Alexander,   F.    The   influence   of   psychologic   factors   upon   gastro- 
intestinal disturbances:   a  symposium.  1.  General  principles,  objectives,  and 
preliminary  results.  Psychoanal.  Quart.,  1934,  3,  501-539. 

5.  Alexander,  F.  Emotional  factors  in  essential  hypertension,  presenta- 
tion of  a  tentative  hypothesis.  Psychosom.  Med.,  1939,  1,  173-179. 

6.  Allport,  G.  W.  Personality.  New  York:  Holt,  1937. 

7.  Allport,  G.  W.  Scientific  models  and  human  morals.  Psychol.  Rev., 
1947,  54,  182-192.  Also  in  [8,  pp.  187-197]. 

8.  Allport,  G.  W,  The  nature  of  personality:  selected  papers.  Reading, 
Mass.:  Addison- Wesley,  1950. 

9.  Allport,  G.  W.  Becoming:  basic  considerations  for  a  psychology  of 
personality.  New  Haven3  Conn.:  Yale  Univer.  Press,  1955. 

10.  Beach,  F.  A.,  &  Jaynes,  J.  Effects  of  early  experience  upon  the  be- 
havior of  animals.  Psychol  Bull,  1954,  51,  239-263. 

11.  Benjamin,  J.  D.  Methodological  considerations  in  the  validation  and 
elaboration  of  psychoanalytical  personality  theory.  Amer.  J.  Orthopsychiat., 
1950,  20,  139-156. 

12.  Bergman,  P.  A  religious  conversion  in  the  course  of  psychotherapy. 
Amer.  J.  Psychother.,  1953,  7,  41-58. 

13.  Bernfeld,  S.  Die  Gestalttheorie.  Imago,  1934,  20,  32-77. 

14.  Bernfeld,  S.  The  facts  of  observation  in  psychoanalysis.  /.  Psychol., 
1941,  12,  289-305. 

15.  Bernfeld,  S.  Freud's  earliest  theories  and  the  school  of  Helmholtz. 
Psychoanal  Quart.,  1941,  13,  341-362. 


16.  Bernfeld,   S.   Freud's  scientific  beginnings.  Amer.  Imago,   1949,   6, 

17.  Bernfeld,  S.  Sigmund  Freud,  M.D.,  1882-1885.  Int.  J.  Psycho-Anal, 
1951,  32,  204-217. 

18.  Bernfeld,  S.,  &  Bemfeld,  S.  G.  Freud's  first  year  in  practice,  1886- 
1887.  Bull  Menninger  Clin.,  1952,  16,  37-49. 

19.  Bemheim,   H.    Die   Suggestion    und    ihre   Heilwirkung.    S.    Freud 
(Trans.).  (2ded.)  Leipzig:  Deuticke,  1896. 

20.  Bertalanffy,   L.   von.   The  theory  of  open  systems  in  physics  and 
biology.  Science,  1950,  111,  23-30. 

21.  Bettelheim,  B.  Individual  and  mass  behavior  in  extreme  situations. 
/.  abnorm.  soc.  Psychol,  1943,  38,  417-452. 

22.  Bettelheim,  B.  Love  is  not  enough:  the  treatment  of  emotionally 
disturbed  children.  Glencoe,  111.:  Free  Press,  1950. 

23.  Bettelheim,  B.  Truants  from  life:  the  rehabilitation  of  emotionally 
disturbed  children.  Glencoe,  111.:  Free  Press,  1955. 

24.  Beveridge,  W.  I.  B.  The  art  of  scientific  investigation.  New  York: 
Norton,  1950. 

25.  Bexton,  W.  H.,  Heron,  W.,  &  Scott,  T.  H.  Effects  of  decreased 
variation  in  the  sensory  environment.  Canad.  J.  PsychoL,  1954,  8,  70-76. 

26.  Bibring,  E.  The  development  and  problems  of  the  theory  of  the 
instincts.  Int.  /.  Psycho-Anal,  1941,  22,  102-131. 

27.  Bibring,  E.  The  mechanism  of  depression.  In  P.  Greenacre  (Ed.), 
Affective  disorders.  New  York:   International  Universities  Press,  1953.  Pp. 

28.  Bibring,    E.    Psychoanalysis    and    the   dynamic   psychotherapies.    /. 
Amer.  psychoanal  Ass.,  1954,  2,  745-770. 

29.  Blake,  R.,  &  Ramsey,  G.   (Eds.)   Perception:  an  approach  to  per- 
sonality. New  York:  Ronald,  1951. 

30.  Bond,  D.  D.  The  love  and  fear  of  flying.  New  York:  International 
Universities  Press,  1952. 

31.  Boring,  E.  G.  A  history  of  experimental  psychology.  (2d  ed.)   New 
York:  Appleton-Century-Crofts,  1950. 

32.  Boring,  E.  G.  Dual  role  of  the  Zeitgeist  in  scientific  creativity.  Sci. 
Monthly,  1955,  80,  101-106. 

33.  Brenman,  M.,  &  Gill,  M.  M.  Hypnotherapy.  New  York:   Interna- 
tional Universities  Press,  1947. 

34.  Brenman,  M.,  &  Gill,  M.  M.  Research  in  psychotherapy  round  table, 
1947.  Amer.  ].  Orthopsychiat.,  1948,  18,  92-118. 

35.  Breuer,  J.,  &  Freud,  S.  Studies  in  hysteria.  New  York:  Nervous  and 
Mental  Disease  Publications,  1937. 

36.  Brown,  J.  F.  Ueber  die  dynamischen  Eigenschaften  der  Realitaets- 
und  Irrealitaetsschichten.  Psychol  Forsch.,  1933,  18,  2-26. 

37.  Bruner,  J.,  &  Goodman,  C.  Value  and  need  as  organizing  factors 
in  perception.  /.  abnorm.  soc.  Psychol,  1947,  42,  33-44. 

38.  Bruner,  J.,  &  Postman,  L.  Emotional  selectivity  in  perception  and 
reaction.  /.  Pers.,  1947,  16,  69-77. 

The  Structure  of  Psychoanalytic  Theory  169 

39.  Bruner,  J.5  &  Postman,  L.  Tension  and  tension  release  as  organizing 
factors  in  perception.  /.  Pers.,  1947,  15,  300-308. 

40.  Bniner,  J.,  &  Postman,  L.  Symbolic  value  as  an  organizing  factor  In 
perception.  /.  soc.  PsychoL,  1948,  27,  203-208. 

41.  Bruner,  J.,  Postman,  L.,  &  McGinnies,  E.  Personal  values  as  selective 
factors  in  perception.  /.  abnorm.  soc.  PsychoL,  1948,  43,  142-154. 

42.  Buehler,  K.  Die  geistige  Entwicklung  des  Kindes.  (6th  ed.)  Jena: 
Fischer,  1930. 

43.  Chein,    I.    The   genetic   factor   in    ahistorical   psychology.   /.    gen. 
PsychoL,  1947,  36,  151-172. 

44.  Delbrueck,  M.  A  physicist  looks  at  biology.  Trans.  Conn.  Acad.  Arts 
ScL,  1949,  38,  175-190. 

45.  Deutsch,  F.  The  choice  of  organ  in  organ  neuroses.  Int.  J.    Psycho- 
Anal,  1939,  20,  252-262. 

46.  Dingle,  H.  The  scientific  adventure.  London:   Pitman,  1952.  New 
York:  Philosophical  Library,  1953. 

47.  Diven,  K.  Certain  determinants  In  the  conditioning  of  anxiety  re- 
actions. /.  PsychoL,  1937,  3,  219-308. 

48.  Bollard,   J.,   &   Miller,  N.   E.   Personality  and  psychotherapy:  an 
analysis  in  terms  of  learning,  thinking  and  culture.  New  York:   McGraw- 
Hill,  1950. 

49.  Dunbar,  H.  F.  Emotions  and  bodily  changes.  (3d  ed.)  New  York: 
Columbia  Univer.  Press,  1947. 

50.  Dunbar,  H.  F.  Mind  and  body:  psychosomatic  medicine.  New  York: 
Random  House,  1947. 

51.  Eissler,  K.  R.  The  effect  of  the  structure  of  the  ego  on  psychoanalytic 
technique.  /.  Amer.  psychoanal.  Ass.,  1953,  1,  104-143. 

52.  Ellis,    W.    D.    A    sourcebook    of    Gestalt   psychology.   New   York: 
Humanities  Press,  1950. 

53.  Erickson,  M.  H.  Experimental  demonstration  of  the  psychopathology 
of  everyday  life.  Psychoanal.  Quart.,  1939,  8,  338-353. 

54.  Erickson,  M.   H.,  &  Kubie,  L.   S.  The  translation  of  the  cryptic 
automatic  writing  of  one  hypnotic  subject  by  another  in  a  trance-like  dis- 
sociated state.  Psychoanal  Quart.,  1940,  9,  51-63. 

55.  Erickson,  M.  H.,  &  Kubie,  L.  S.  The  successful  treatment  of  a  case 
of  acute  hysterical  depression  by  a  return  under  hypnosis  to  a  critical  phase 
of  childhood.  Psychoanal  Quart.,  1941,  10,  583-609. 

56.  Erikson,  E.  H.  Configurations  in  play — clinical  notes.  Psychoanal. 
Quart.,  1937,  6,  139-214. 

57.  Erikson,  E.  H.  Observations  on  Sioux  education.  /.  PsychoL,  1939, 
7,  101-156. 

58.  Erikson,    E.    H.    Problems    of    infancy    and    early    childhood.    In 
Cyclopedia  of  medicine.  Philadelphia:   Davis,  1940.  Pp.  714-730.  Also  in 
G.  Murphy  &  A.  Bachrach  (Eds.),  Outline  of  abnormal  psychology.  New 
Fork:  Modern  Library,  1954.  Pp.  3-36. 

59.  Erikson,  E.  H.  Childhood  and  tradition  in  two  American  Indian 
tribes.  In  The  psychoanalytic  study  of  the  child.  Vol.  1.  New  York:  Inter- 


national  Universities  Press,  1945.  Pp.  319-350.  Also  (revised)  in  C.  Kluck- 
hohn  &  H.  Murray  (Eds/),  Personality  in  nature,  society  and  culture.  New 
York:  Knopf,  1948.  Pp.  176-203. 

60.  Erikson,  E.  H.  Ego  development  and  historical  change.  In  The  psy- 
choanalytic study   of   the   child.   Vol.   2.   New  York:    International   Uni- 
versities Press,  1946.  Pp.  359-396. 

61.  Erikson,  E.  H.  Childhood  and  society.  New  York:  Norton,  1950. 

62.  Erikson,  E.  H.  Growth  and  crises  of  the  "healthy  personality."  In  C. 
Kluckhohn  &  H.  Murray  (Eds.) ,  Personality  in  nature,  society  and  culture. 
(2d  ed.)  New  York:  Knopf,  1953.  Pp.  185-225. 

63.  Erikson,   E.  H.  The  dream  specimen  of  psychoanalysis.  /.  Amer. 
psychoanal  Ass.,  1954,  2,  5-56.  Also  in  R.  P.  Knight  £  G.  R.  Friedman 
(Eds.),  Psychoanalytic  psychiatry  and  psychology,  clinical  and  theoretical 
papers.  Austen  Riggs  Center.  Vol.  1.  New  York:  International  Universities 
Press,  1954.  Pp.  131-170. 

64.  Erikson,  E.   H.    Freud's   "The  origins  of  psychoanalysis."  Int.   J. 
Psycho-Anal,  1955,  36,  1-15. 

65.  Erikson,  E.  H.  The  first  psychoanalyst:    crisis  and  discovery.   Yale 
Rev.,  1956,  46,  40-62. 

66.  Erikson,  E.  H.  The  problem  of  ego  identity.  /.  Amer.  psychoanal. 
Ass.,  1956,  4,  56-121. 

67.  Escalona,  S.  Problems  in  psycho-analytic  research.  Int.  J.  Psycho- 
Anal,  1952,32,  11-21. 

68.  Farber,  L.,  &  Fisher,  C.  An  experimental  approach  to  dream  psy- 
chology through  the  use  of  hypnosis.  Psychoanal  Quart.,  1943,  12,  202-216. 

69.  Fechner,  G.  T.  Einige  Ideen  zur  Schopfungs-  und  Entwicklungsge- 
schichte  der  Organismen.  Leipzig:  Breitkopf  und  Hartel,  1873. 

70.  Federn,   P.  Ego  psychology  and  the  psychoses.  New  York:    Basic 
Books,  1952. 

71.  Feigl,  H.   Principles  and  problems  of  theory  construction  in  psy- 
chology.  In  Current  trends  in  psychological  theory.   Pittsburgh:    Univer. 
Pittsburgh  Press,  1951.  Pp.  179-213. 

72.  Fenichel,    O.    Problems   of   psychoanalytic    technique.    New  York: 
Psychoanalytic  Quarterly,  Inc.,  1941. 

73.  Fenichel,  O.   The  psychoanalytic   theory  of  neurosis.  New  York: 
Norton,  1945. 

74.  Fenichel,  O.  A  critique  of  the  death  instinct.  In  [77,  first  series, 
pp.  363-372]. 

75.  Fenichel,  O.  The  ego  and  the  affects.  In  [77,  second  series,  pp. 

76.  Fenichel,   O.   Psychoanalytic  remarks   on   Fromm's  book,   "Escape 
from  freedom."  In  [77,  second  series,  pp.  260-277]. 

77.  Fenichel,  O.  Collected  papers.  First  and  second  series.  New  York: 
Norton,  1953-54. 

78.  Ferenczi,  S.  Stages  in  the  development  of  the  sense  of  reality.  In 
Sex  in  psychoanalysis.  New  York:  Brunner,  1950.  Pp.  213-239. 

The  Structure  of  Psychoanalytic  Theory  171 

79.  Ferenczi,   S.   The  problem  of  acceptance  of  unpleasant  ideas.   In 
Further  contributions  to  the  theory  and  technique  of  psycho-analysis.  Lon- 
don: Hogarth,  1950.  Pp.  366-379. 

80.  Ferenczi,  S.  The  psyche  as  an  inhibiting  organ.  In  Further  contribu- 
tions to   the  theory  and  technique  of  psycho-analysis.  London:    Hogarth, 
1950.  Pp.  379-383. 

81.  Ferenczl,  S.  Thinking  and  muscle  innervation.  In  Further  contribu- 
tions to   the  theory  and  technique  of  psycho-analysis.  London:    Hogarth, 
1950.  Pp.  230-232. 

82.  Finch,  G..  &  Culler,  E.  Relation  of  forgetting  to  experimental  ex- 
tinction. Amer.  J.  PsychoL,  1935,  47,  656-662. 

83.  Fisher,  G.  Studies  on  the  nature  of  suggestion.  Parts  I  and  II.  /. 
Amer.  psychoanal.  Ass.,  1953,  1,  222-255,  406-437. 

84.  Fisher,   G.  Dreams  and  perception:    the  role  of  preconscious  and 
primary  modes  of  perception  in  dream  formation.  /.  Amer.  psychoanal.  Ass., 
1954,  2,  389-445. 

85.  French,  T.   M.   Interrelations  between  psychoanalysis  and  the  ex- 
perimental work  of  Pavlov.  Amer.  J.  Psychiat.,  1933,  12  (89),  1165-1203. 

86.  French,  T.  M.  Psychogenic  factors  in  asthma.  Amer.  J.  Psychiat.., 
1939,  96,  87-101. 

87.  French,  T.  M.  Goal,  mechanism  and  the  integrative  field.  Psychosom. 
Med.,  1941,  3,  226-252. 

88.  French,  T.  M.  The  integrative  process.  Dialectic  a,  1951,  5,  246-256. 

89.  French,  T.  M.  The  integration  of  behavior.  Vols.  1  &  2.  Chicago: 
Univer.  Chicago  Press,  1952,  1954. 

90.  French,  T.  M.,  &  Alexander,  F.  Psychogenic  factors  in  bronchial 
asthma.  Parts  I  and  II.  Psychosom.  Med.  Monogr.,  1941,  1  (4);  2  (1,  2). 

91.  Frenkel-Brunswik,    E.    Motivation    and   behavior.    Genet.   Psychol. 
Monogr.,  1942,  26,  121-265. 

92.  Frenkel-Brunswik,  E.  Psychoanalysis  and  the  unity  of  science.  Proc. 
Amer.  Acad.  Arts.  ScL,  1954,  80,  271-350. 

93.  Freud,  A.    (1936)    The  ego  and  the  mechanisms  of  defence.  New 
York:  International  Universities  Press,  1946. 

94.  Freud,   S.    (1887-1902)    The   origins  of  psycho-analysis:  letters  to 
Wilhelm  Fliess,  drafts  and  notes:  1887-1902.  New  York:  Basic  Books,  1954. 

95.  Freud,  S.  (1893-1914)   Collected  papers  of  .  .  .  .  Vol.  1.  London: 
Hogarth,  1948. 

96.  Freud,  S.  (1894)  The  defence  neuro-psychoses.  In  [95,  pp.  59-75]. 

97.  Freud,  S.   (1896)   Further  remarks  on  the  defence  neuro-psychoses. 
In  [95,  pp.  155-182]. 

98.  Freud,  S.   (1900)   The  interpretation  of  dreams.  In  The  basic  writ- 
ings of  ....  New  York:  Modern  Library,  1938.  Pp.  179-549. 

99.  Freud,   S.    (1904)    The  psychopathology  of  everyday  life.  In  The 
basic  writings  of  .  .  .  .  New  York:  Modern  Library,  1938.  Pp.  33-178. 

100.  Freud,  S.  (1905)  Wit  and  its  relation  to  the  unconscious.  In  The 
basic  writings  of  .  .  .  .  New  York:  Modern  Library,  1938.  Pp.  631-803. 


101.  Freud,  S.  (1905)  Three  essays  on  the  theory  of  sexuality.  London: 
Imago,  1949. 

102.  Freud,  S.   (1907)  Delusion  and  dream.  London:  Allen  &  Unwin, 

103.  Freud,  S.   (1908)    The  relation  of  the  poet  to  day-dreaming.  In 
Collected  papers  of  .  .  .  .  Vol.  4.  London:  Hogarth,  1948.  Pp.  173-183. 

104.  Freud,  S.   (1910)   The  antithetical  sense  of  primal  words.  In  Col- 
lected papers  of  .  .  .  .  Vol.  4.  London:  Hogarth,  1948.  Pp.  184-191. 

105.  Freud,  S.  (1910)  Leonardo  da  Vinci.  New  York:  Random  House, 

106.  Freud,  S.   (1910-1919)    Papers  on  technique.  In  Collected  papers 
of  .  .  .  .  Vol.  2.  London:  Hogarth,  1948.  Pp.  285-402. 

107.  Freud,  S.   (1911)   Psycho-analytic  notes  upon  an  autobiographical 
account  of  a  case  of  paranoia  (dementia  paranoides).  In  Collected  papers 
of  .  .  .  .  Vol.  3.  London:  Hogarth,  1948.  Pp.  387-470. 

108.  Freud,   S.    (1911)    Formulations  regarding  the  two  principles  in 
mental  functioning.  In  Collected  papers  of  .  .  .  .  Vol.  4.  London:  Hogarth, 
1948.  Pp.  13-21. 

109.  Freud,   S.    (1912)    Totem,  and  taboo.    In   The  basic   writings  of 
....  New  York:  Modern  Library,  1938.  Pp.  805-930. 

110.  Freud,  S.  (1912)  A  note  on  the  unconscious  in  psycho-analysis.  In 
Collected  papers  of  ....  Vol.  4.  London:  Hogarth,  1948.  Pp.  22-29. 

111.  Freud,  S.  (1913)  The  occurrence  in  dreams  of  material  from  fairy- 
tales. In  Collected  papers  of  .  .  .  .  Vol.  4.  London:   Hogarth,   1948.  Pp. 

112.  Freud,  S.    (1913)    The  theme  of  the  three  caskets.  In  Collected 
papers  of  .  .  .  .  Vol.  4.  London:  Hogarth,  1948.  Pp.  244-256. 

113.  Freud,  S.  (1914)  The  Moses  of  Michelangelo.  In  Collected  papers 
of  .  .  .  .  Vol.  4.  London:  Hogarth,  1948.  Pp.  257-287. 

114.  Freud,  S.    (1914)    On  narcissism:    an  introduction.  In  Collected 
papers  of  .  .  .  .  Vol.  4.  London:  Hogarth,  1948.  Pp.  30-59. 

115.  Freud,   S.    (1915)    Instincts  and   their   vicissitudes.    In    Collected 
papers  of  .  .  .  .  Vol.  4.  London:  Hogarth,  1948.  Pp.  60-83. 

116.  Freud,  S.  (1915)  Repression.  In  Collected  papers  of  .  .  .  .  Vol.  4. 
London:  Hogarth,  1948.  Pp.  84-97. 

117.  Freud,  S.   (1915)   The  unconscious.  In  Collected  papers  of  .  .  .  . 
Vol.  4.  London:  Hogarth,  1948.  Pp.  98-136. 

118.  Freud,  S.  (1916)  A  mythological  parallel  to  a  visual  obsession.  In 
Collected  papers  of  .  .  .  .  Vol.  4.  London:  Hogarth,  1948.  Pp.  345-346. 

119.  Freud,  S.   (1916)   Metapsychological  supplement  to  the  theory  of 
dreams.  In  Collected  papers  of  .  .  .  .  Vol.  4.  London:  Hogarth,  1948.  Pp. 

120.  Freud,  S.  (1917)  Mourning  and  melancholia.  In  Collected  papers 
of  .  .  .  .  Vol.  4.  London:  Hogarth,  1948.  Pp.  152-170. 

121.  Freud,  S.   (1917)   A  general  introduction  to  psycho-analysis.  New 
York:  Perma  Giants,  1949. 

The  Structure  of  Psychoanalytic  Theory  173 

122.  Freud,  S.    (1919).  A  cMId  is  being  beaten.  In  Collected  papers 
of  .  .  .  .  Vol.  2.  London:  Hogarth,  1948.  Pp.  172-201. 

123.  Freud,  S.  (1920)  Beyond  the  pleasure  principle.  London:  Hogarth. 

124.  Freud,  S.   (1921)    Group  psychology  and  the  analysis  of  the  ego. 
London:  Hogarth,  1948. 

125.  Freud,  S.    (1922)    Medusa's  head.  In  Collected  papers  of  .  .  .  . 
Vol.  5.  London:  Hogarth,  1950.  Pp.  105-106. 

126.  Freud,  S.  (1923)  The  ego  and  the  id.  London:  Hogarth,  1947. 

127.  Freud,  S.   (1924)   Neurosis  and  psychosis.  In  Collected  papers  of 
....  Vol.  2.  London:  Hogarth,  1948.  Pp.  250-254. 

128.  Freud,  S,  (1924)  The  loss  of  reality  in  neurosis  and  psychosis.  In 
Collected  papers  of  .  .  .  .  Vol.  2.  London:  Hogarth,  1948.  Pp.  277-282. 

129.  Freud,  S.  (1925)   Negation.  In  Collected  papers  of  .  .  .  .  Vol.  5. 
London:  Hogarth,  1950.  Pp.  181-185.  Also  in  [268,  pp.  338-348]. 

130.  Freud,  S.   (1925)   A  note  upon  the  "mystic  writing-pad.55  In  Col- 
lected papers  of  .  .  .  .  Vol.  5.  London:  Hogarth,  1950.  Pp.  175-180.  Also 
in  [268,  pp.  329-337]. 

131.  Freud,   S.    (1926)    The  problem  of  anxiety.  New  York:    Psycho- 
analytic Quarterly  and  Norton,  1936. 

132.  Freud,  S.  (1927)  The  future  of  an  illusion.  New  York:  Liveright, 

133.  Freud,  S.  (1928)  Dostoevsky  and  parricide.  In  Collected  papers  of 
....  Vol.  5.  London:  Hogarth,  1950.  Pp.  222-242. 

134.  Freud,  S.  (1928)   Humour.  In  Collected  papers  of  .  .  .  .  Vol.  5. 
London:  Hogarth,  1950.  Pp.  215-221. 

135.  Freud,  S.    (1928)    A  religious  experience.  In  Collected  papers  of 
....  Vol.  5.  London:  Hogarth,  1950.  Pp.  243-246. 

136.  Freud,  S.  (1930)  Civilization  and  its  discontents.  London:  Hogarth, 

137.  Freud,   S.    (1932)    New   introductory  lectures  on  psycho-analysis. 
New  York:  Norton,  1933. 

138.  Freud,  S.  (1932)  Why  war?  In  John  Rickman  (Ed.),  Civilization, 
war,  and  death.  London:  Hogarth,  1939.  Pp.  82-97. 

139.  Freud,  S.    (1937)    Analysis  terminable  and  interminable.  In  Col- 
lected papers  of  .  .  .  .  Vol.  5.  London:  Hogarth,  1950.  Pp.  316-357. 

140.  Freud,  S.  (1938)  An  outline  of  psychoanalysis.  New  York:  Norton, 

141.  Freud,  S.  (1939)  Moses  and  monotheism.  London:  Hogarth,  1939. 

142.  Fromm-Reichmann,  F.  Transference  problems  in  schizophrenics. 
Psychoanal.  Quart.,  1939,  8,  412-426. 

143.  Fromm-Reichmann,  F.  Notes  on  the  development  of  treatment  of 
schizophrenics  by  psychoanalytic  psychotherapy.  Psychiatry,  1948,  11,  263- 

144.  Gardner,  R.  W.  Cognitive  styles  in  categorizing  behavior.  /.  Pers.y 
1953,  22,  214^233. 


145.  GUI,  M.  M.  Psychoanalysis  and  exploratory  psychotherapy.  /.  Amer. 
psychoanal  Ass.,  1954,  2?  771-797. 

146.  Gill,  M.  M.  The  present  state  of  psychoanalytic  theory.  /.  abnorm. 
soc.  PsychoL,  1959,  58,  1-8. 

147.  Gill,  M.  M.,  &  Brenman,  M.  Problems  in  clinical  research  round 
table,  1946.  Amer.  ].  OrthopsychiaL,  1947,  17,  196-230. 

148.  Gill,  M.  M.,  &  Brenman,  M.  Hypnosis  and  related  states:  psycho- 
analytic studies  in  regression.  New  York:   International  Universities  Press, 
in  press. 

149.  Gill,  M.  M.,  &  Rapaport,  D.  The  points  of  view  and  assumptions 
of  metapsychology.  Int.  J.  Psycho-Anal.,  in  press. 

150.  Glover,  E.  Basic  mental  concepts.  London:  Imago,  1947. 

151.  Glover,   E.   The   technique  of  psycho-analysis.    (Rev.   Ed.)    New 
York:  International  Universities  Press,  1955. 

152.  Goldmann,  A.  E.  Studies  in  vicariousness :  degree  of  motor  activity 
and    the    autokinetic    phenomenon.    Amer.    ].    PsychoL,    1953,    66,    613- 

153.  Gross,  A.  The  Secret  Bull  Menninger  Clin.3  1951,  15,  37-44. 

154.  Gross,  A.  Freud3s  creative  period.  Unpublished  manuscript. 

155.  Halverson,  H.  M.  Infant  sucking  and  tensional  behavior.  /.  genet. 
PsychoL,  1938,  53,  365-430. 

156.  Halverson,   H.   M.   Genital   and  sphincter  behavior  of  the  male 
infant  /.  genet.  PsychoL,  1940,  56,  95-136. 

157.  Hartmann,   H.   Ego  psychology  and  the  problem  of  adaptation, 
New  York:   International  Universities  Press,  1958.  Also,  abridged,  in  [268, 
pp.  362-396]. 

158.  Hartmann,  H.  Psychoanalysis  and  the  concept  of  health.  Int.  J. 
Psycho- Anal.,  1939,  20,  308-321. 

159.  Hartmann,  H.  On  rational  and  irrational  action.  In  G.  Roheim 
(Ed.),  Psychoanalysis  and  the  social  sciences.  Vol.   1.  New  York:    Inter- 
national Universities  Press,  1947.  Pp.  359-392. 

160.  Hartmann,   H.    Comments   on   the   psychoanalytic   theory   of   in- 
stinctual drives.  Psychoanal.  Quart.,  1948,  17,  368-388. 

161.  Hartmann,  H.  Comments  on  the  psychoanalytic  theory  of  the  ego. 
In  The  psychoanalytic  study  of  the  child.  Vol.  5.  New  York:  International 
Universities  Press,  1950.  Pp.  74-96. 

162.  Hartmann,  H.  The  mutual  influences  in  the  development  of  the 
ego  and  id.  In  The  psychoanalytic  study  of  the  child.  Vol.  7.  New  York: 
International  Universities  Press,  1952.  Pp.  9-30. 

163.  Hartmann,  H.  Contribution  to  the  metapsychology  of  schizophrenia. 
In  The  psychoanalytic  study  of  the  child.  Vol.  8.  New  York:  International 
Universities  Press,  1953.  Pp.  177-198. 

164.  Hartmann,  H.  Notes  on  the  theory  of  sublimation.  In  The  psy- 
choanalytic study  of  the  child.  Vol.    10.  New  York:    International  Uni- 
versities Press,  1955.  Pp.  9-29. 

165.  Hartmann,  H.  The  development  of  the  ego  concept  in  Freud's 
work.  Int.  J.  Psycho- Anal,  1956,  37,  425-438. 

The  Structure  of  Psychoanalytic  Theory  175 

166.  Hartmann,  H.3  &  Kris,  E.  The  genetic  approach  in  psychoanalysis. 
In  The  psychoanalytic  study  of  the  child.  Vol.  L  New  York:  International 
Universities  Press,  1945.  Pp.'  11-29. 

167.  Hartmann,  H.,  Kris,  E..  &  Loewenstein,  R.  M.  Comments  on  the 
formation  of  psychic  structure.  In  The  psychoanalytic  study  of  the  child. 
Vol.  2.  New  York:  International  Universities  Press,  1946.  Pp.  11-38. 

168.  Hartmann.  H.,  Kris,  E.,  &  Loewenstein,  R.  M.  Notes  on  the  theory 
of  aggression.  In  The  psychoanalytic  study  of  the  child.  Vols.  3  &  4.  New 
York:  International  Universities  Press,  1949.  Pp.  9-36. 

169.  Hebb,  D.  O.  The  organization  of  behavior:  a  neuro  psychological 
theory.  New  York:  Wiley,  1949. 

170.  Heron,  W.  The  pathology  of  boredom.  Sci.  Amer.,  1957,  196,  52-56. 

171.  Heron,  W.,  Bexton,  W.  H.,  &  Hebb,  D.  O.  Cognitive  effects  of  a 
decreased  variation  in  the  sensory  environment.  Amer.  Psychol.,  1953,  8, 

172.  Heron,  W.,  Doane,  B.  K.,  &  Scott,  T.  H.  Visual  disturbances  after 
prolonged  perceptual  isolation.  Canad.  /.  PsychoL,  1956,  10,  13-18. 

173.  Hilgard,    E.    R.    Experimental   approaches    to   psychoanalysis.   In 
E.    Pumpian-Mindlin    (Ed.),   Psychoanalysis  as  science.   Stanford,   Calif.: 
Stanford  Univer.  Press,  1952.  Pp.  3-45. 

174.  Holt,  E.  B.  The  Freudian  wish  and  its  place  in  ethics.  New  York: 
Holt,  1915. 

175.  Holt,  E.  B.  Animal  drive  and  the  learning  process.  New  York: 
Holt,  1931. 

1 76.  Holt,  R.  R.  Implications  of  some  contemporary  personality  theories 
for  Rorschach  rationale.  In  Klopfer,  B.,  Ainsworth,  M.,  Klopfer,  W.,  &  Holt, 
R.  R.,  Developments  in  the  Rorschach  technique.  Vol.  1.  Yonkers,  N.Y.: 
World  Book  Co.,  1954.  Pp.  501-560. 

177.  Holt,  R.  R.  Freud  and  the  scientific  method  of  psychoanalysis.  Un- 
published manuscript. 

178.  Holt,  R.  R.  Gauging  primary  and  secondary  processes  in  Rorschach 
responses.  /.  pro].  Tech.,  1956,  20,  14-25. 

179.  Holzman,  P.  S.,  &  Klein,  G.  S.  Cognitive  system-principles  of  level- 
ing and  sharpening:   individual  differences  in  assimilation  effects  in  visual 
time-error.  /.  PsychoL,  1954,  37,  105-122. 

180.  Holzman,  P.  S.  The  relation  of  assimilation  tendencies  in  visual, 
auditory,  and  kinesthetic  time-error  to  cognitive  attitudes  of  leveling  and 
sharpening.  /.  Pers.3  1954,  22,  375-394. 

181.  Homey,  K.   The  neurotic  personality  of  our  time.  New  York: 
Norton,  1937. 

182.  Homey,  K.  New  ways  in  psychoanalysis.  New  York:  Norton,  1939. 

183.  Hull,  C.  L.  Psychology  seminar  memoranda,  1934-1937.  Unpub- 
lished manuscript,  on  deposit  at  Yale  Univer.  Library. 

184.  Humphrey,  G.  The  conditioned  reflex  and  the  Freudian  wish.  /. 
abnorm.  soc.  PsychoL,  1920,  14,  388-392. 

185.  Humphrey,    G.    Education    and    Freudianism.    /.    abnorm.    soc. 
PsychoL,  1921,  15,  350-402. 


186.  Hunt,  J.  McV.  The  effects  of  infant  feeding-frustration  upon  adult 
hoarding  in  the  albino  rat.  /.  abnorm.  soc,  PsychoL,  1941,  36,  338-360. 

187.  Hunt,  J.  McV.,  Schlosberg,  H.,  Solomon,  R.  L.,  &  Stellar,  E. 
Studies  of  the  effect  of  infantile  experience  on  adult  behavior  in  rats.  I. 
Effects  of  infantile  feeding  frustration  on  adult  hoarding.  /.  comp.  physiol. 
PsychoL,  1947,  40,  291-304. 

188.  Huston,  P.  E.,  Shakow,  D.,  &  Erickson,  M.  H.  A  study  of  hypnot- 
ically induced  complexes  by  means  of  the  Luria  technique.  /.  gen.  PsychoL, 
1934,  11,  65-97. 

189.  Jacobson,  E.  Observations  on  the  psychological  effect  of  imprison- 
ment on  female  political  prisoners.  In  K.   Eissler   (Ed.),  Searchlights  on 
delinquency.  New  York:  International  Universities  Press,  1949.  Pp.  341-368. 

190.  Jacobson,  E.   Contribution  to  the  metapsychology  of  cyclothymic 
depression.  In  P.  Greenacre  (Ed.),  Affective  disorders.  New  York:   Inter- 
national Universities  Press,  1953.  Pp.  49-83. 

191.  Jacobson,  E.  The  affects  and  their  pleasure-unpleasure  qualities, 
in  relation  to  psychic  discharge  processes.  In  R.  Loewenstein  (Ed.),  Drives, 
affects,  behavior.  New  York:    International  Universities  Press,   1953.   Pp., 

192.  Jacobson,  E.  The  self  and  the  object  world.  In  The  psychoanalytic 
study  of  the  child.  Vol.  9.  New  York:  International  Universities  Press,  1954. 
Pp.  75-127. 

193.  Jones,  E.  The  life  and  work  of  Sigmund  Freud,  Vols.  1  &  2.  New 
York:  Basic  Books,  1953,  1955. 

194.  Kardiner,  A.  The  individual  and  his  society:  the  psycho  dynamics 
of  primitive  social  organization.  New  York:  Columbia  Univer.  Press,  1939. 

195.  Kardiner,  A.  The  psychological  frontiers  of  society.  New  York: 
Columbia  Univer.  Press,  1945. 

196.  Kempf,  E.  J.  The  autonomic  functions  and  the  personality.  New 
York:  Nervous  and  Mental  Disease  Publications,  1921. 

197.  Klein,  G.  S.  Need  and  regulation.  In  M.  Jones  (Ed.),  Nebraska 
symposium  on  motivation.  Lincoln,  Neb.:   Univer.  Nebraska  Press,   1954. 
Pp.  224-274. 

198.  Klein,  G.  S.  Perception,  motive,  and  personality:  a  clinical  perspec- 
tive. To  be  published  by  Basic  Books.  See  also  in  J.  McCary  (Ed.),  Psy- 
chology of  personality:  six  modern  approaches.  New  York:    Logos  Press, 
1956.  Pp.  123-199. 

199.  Klein,  G.  S.,  &  Schlesinger,  H.  J.  Perceptual  attitudes  toward  in- 
stability. I.  Prediction  of  apparent  movement  experiences  from  Rorschach 
responses.  /.  Pers.,  1951,  19,  289-302. 

200.  Klein,  G.  S.,  Schlesinger,  H.  J.,  &  Meister,  D.  E.  The  effects  of 
personal  values  on  perception:    an  experimental  critique.  PsychoL  Rev,, 
1951,  58,  96-112. 

201.  Klein,  G.  S.,  Spence,  D.  P.,  Holt,  R.  R.,  &  Gourevitch,  S.  Pre- 
conscious   influences  upon   conscious   cognitive  behavior.   Amer.   PsychoL, 
19553  10,  380.   (Abstract)    Published  as  Cognition  without  awareness:    I. 

The  Structure  of  Psychoanalytic  Theory  177 

Subliminal   Influences   upon   conscious   thought.   /.   abnorm.   soc.   PsychoL, 
1958.  57,  255-266. 

202.  Knight,  R.  P.  Borderline  States.  Bull  Menninger  Clin.}  1953,  17, 
1-12.  Also  In  R.  P.  Knight  &  G.  R.  Friedman   fEds.K  Psychoanalytic  psy- 
chiatry and  psychology,  clinical  and  theoretical  papers.  Austen  Riggs  Center. 
Vol.  1.  New  York:   International  Universities  Press,  1954.  Pp.  97-109. 

203.  Knight,  R.  P.  Management  and  psychotherapy  of  the  borderline 
schizophrenic  patient.  Bull  Menninger  Clin.,   1953.   17,   139-150.  Also  in 
R.  P.  Knight  &  C.  R.  Friedman  (Eds.),  Psychoanalytic  psychiatry  and  psy- 
chology, clinical  and  theoretical  papers.  Austen  Riggs  Center.  Vol.  1.  New 
York:  International  Universities  Press,  1954.  Pp.  110-122. 

204.  Koffka,  K.  Principles  of  Gestalt  psychology.  New  York:  Harcourt, 
Brace,  1935. 

205.  Kris,  E.  Notes  on  the  development  and  on  some  current  problems 
of  psychoanalytic  child  psychology.  In  The  psychoanalytic  study  of  the  child. 
Vol.  5.  New  York:   International  Universities  Press,  1950.  Pp.  24-46. 

206.  Kris,   E.   On   preconscious  mental   processes.   Psychoanal   Quart., 
1950,  19,  540-560.  Also  in  [268,  pp.  474-493:  and  207,  pp.  303-318]. 

207.  Kris,  E.  Psychoanalytic  explorations  in  art.  New  York:  International 
Universities  Press,  1952. 

208.  Knis,  D.,  Werner,  H.,  &  Wapner.  S.  Studies  in  vlcariousness :  motor 
activity  and  perceived  movement.  Amer.  ].  Psychol,  1953,,  66,,  603-608. 

209.  Kubie,  L.  S.  Problems  in  clinical  research  round  table,  1946.  Amer. 
J.  Orthopsychiat.,  1947,  17,  196-230. 

210.  Kubie,  L.  S.  Problems  and  techniques  of  psychoanalytic  validation 
and  progress.   In  E.   Pumpian-Mindlin    (Ed.),  Psychoanalysis  as  science. 
Stanford,  Calif.:  Stanford  Univer.  Press,  1952.  Pp.  46-124. 

211.  Leeper,    R.    W.    A    motivational    theory    of    emotion    to    replace 
"emotion  as  disorganized  response."  Psychol.  Rev.,  1948,  55,  5-21. 

212.  Levy,  D.  M.  Experiments  on  the  sucking  reflex  and  social  behavior 
of  dogs.  Amer.  J.  Orthopsychiat.,  1934,  4,  203-224. 

213.  Levy,  D.  M.  On  instinct-satiation:  an  experiment  on  pecking  be- 
havior of  chickens.  /.  gen.  Psychol,  1938,  18,  327-348. 

214.  Lewin,  B.  The  psychoanalysis  of  elation.  New  York:  Norton,  1950. 

215.  Lewin,  K.  Gesetz  und  Experiment  in  der  Psychologic.  Symposion, 
1927,  1,  375-421. 

216.  Lewin,  K.  A  dynamic  theory  of  personality.  New  York:  McGraw- 
Hill,  1935. 

217.  Lewin,  K.  Principles  of  topological  psychology.  New  York:   Mc- 
Graw-Hill, 1936. 

218.  Lewin,  K.  Psychoanalysis  and  topological  psychology.  Bull  Men- 
ninger Clin.,  1937,  1,  202-212. 

219.  Lewin,   K.   The   conceptual   representation   and   measurement   of 
psychological  forces.  Contr.  psychol  Theor.,  1,  (4).  Durham,  N.C.:  Duke 
Univer.  Press,  1938. 

220.  Lewin,  K.  Comments  concerning  psychological  forces  and  energies, 
and  the  structure  of  the  psyche.  In  [268,  pp.  76-94]. 


221.  Lewin,  K.  Intention,  will,  and  need.  In  [268,  pp.  95-153]. 

222.  Lilly,   J.    C.   Discussion   in   Illustrative  strategies  for   research   on 
psycho  pathology  in  mental  health.  Group   for  the  Advancement  of  Psy- 
chiatry, Symposium  No.  2,  June,  1956.  Pp.  13-20. 

223.  Lilly,  J.  C.  Mental  effects  of  reduction  of  ordinary  levels  of  physical 
stimuli  on  intact,  healthy  persons.  Psychiat.  Res.  Reports,  1956,  5,  1-9. 

224.  Lorand,  S.  Technique  of  psychoanalytic  therapy.  New  York:   In- 
ternational Universities  Press,  1946. 

225.  Lorenz,  K.  Der  Kumpan  in  der  Umwelt  des  Vogels.  Der  Art- 
genosse  als  ausloesendes  Moment  sozialer  Verbal tungsweisen.  /.   Ornith., 
1935,  83,  137-213. 

226.  McClelland,  D.  C.  The  psychology  of  mental  content  reconsidered. 
Psychol  Rev.,  1955,  62,  297-302. 

227.  MacCorquodale,  K.,  &  Meehl,  P.  E.  Operational  validity  of  inter- 
vening constructs.  In  M.  Marx   (Ed.),  Psychological  theory.  New  York: 
Macmillan,  1951.  Pp.  103-111. 

228.  Maier,  N.  R.  F.  Frustration:  the  study  of  behavior  without  a  goal. 
New  York:  McGraw-Hill,  1949. 

229.  Masserman,  J.  H.  Behavior  and  neurosis:  an  experimental  psycho- 
analytic approach  to  psycho  biologic  principles.  Chicago:   Univer.  Chicago 
Press,  1943. 

230.  Meng,  H.  Das  Problem  der  Organpyschose.  Int.  Z.  PsychoanaL, 

1934,  20,  439-458. 

231.  Menkes,  A.,  &  Menkes,  J.  The  application  of  dimensional  analysis 
to  learning  theory.  Psychol.  Rev.,  1957,  64,  8-13. 

232.  Merlan,  P.  Brentano  and  Freud.  /.  Hist.  Ideas,  1945,  6,  375-377. 

233.  Merlan,  P.  Brentano  and  Freud:  a  sequel.  /.  Hist.  Ideas,  1949,  10, 

234.  Miller,  N.  E.  Studies  of  fear  as  an  acquirable  drive:   I.  Fear  as 
motivation  and  fear-reduction  as  reinforcement  in  the  learning  of  new 
responses.  /.  exp.  Psychol,  1948,  38,  89-101. 

235.  Miller,  N.  E.  Theory  and  experiment  relating  psychoanalytic  dis- 
placement to  stimulus-response   generalization.  /.   abnorm.   soc.   Psychol., 
1948,  43,  155-178. 

236.  Miller,  N.  E.  Comments  on  theoretical  models,  illustrated  by  the 
development  of  a  theory  of  conflict  behavior.  /.  Pers.,  1951,  20,  82-100. 

237.  Miller,  N.  E.,  &  Bollard,  J.  Social  learning  and  imitation.  New 
Haven,  Conn.:  Yale  Univer.  Press,  1941. 

238.  Morgan,  C.  D.,  &  Murray,  H.   A.  A  method  for  investigating 
phantasies:    the   Thematic   Apperception    Test.   Arch.   Neurol.   Psychiat., 

1935,  34,  289-306. 

239.  Mowrer,  O.  H.  Learning  theory  and  personality  dynamics.  New 
York:  Ronald,  1950. 

240.  Munroe,  R.  L.  Schools  of  psychoanalytic  thought.  New  York: 
Dryden,  1955. 

241.  Murphy,  G.  Personality:  a  biosocial  approach  to  origins  and  struc- 
ture. New  York:  Harper,  1947. 

The  Structure  of  Psychoanalytic  Theory  179 

242.  Murray,  H.  A.  The  effect  of  fear  upon  estimates  of  the  malicious- 
ness of  other  personalities.  /.  sac.  Psycho!.,  1933.  4,  310-329. 

243.  Murray,  H.  A.,  et  al  Explorations  in  personality.  New  York:  Ox- 
ford Univer.  Press,  1938. 

244.  Nachmansohn,  M.  Concerning  experimentally  produced   dreams. 
In  [268,  pp.  257-287]. 

245.  Nunberg,  H.  The  synthetic  function  of  the  ego.  In  Practice  and 
theory  of  psychoanalysis.  New  York:  Nervous  and  Mental  Disease  Publica- 
tions, 1948.  Pp.  120-136. 

246.  Page,  J.,  &  Warkentin,  J.  Masculinity  and  paranoia.  /.  abnorm. 
soc.  Psychol,  1938,  33,  527-531. 

247.  Piaget,  J.  Judgment  and  reasoning  in  the  child.  New  York:  Har- 
court,  Brace,  1928. 

248.  Piaget,  J.  The  child's  conception  of  the  world.  New  York:  Har- 
court,  Brace,  1929. 

249.  Piaget,  J.  The  child's  conception  of  physical  causality.  London: 
Routledge,  1930. 

250.  Piaget,  J.  Children's  philosophies.  In  C.  Murchison  (Ed.),  A  hand- 
book  of  child  psychology.  Worcester,  Mass.:    Clark  Univer.  Press,   1931. 
Pp.  377-391. 

251.  Piaget,   J.    The   language   and   thought   of   the   child.    (2d.   ed.) 
London:  Routledge,  1932. 

252.  Piaget,  J.  Introduction  a  Fepistemologie  genetique.  Vols.  1,  2,  &  3. 
Paris:  Presses  Universitaires  de  France,  1950. 

253.  Piaget,  J.  The  psychology  of  intelligence.  New  York:   Harcourt, 
Brace,  1950. 

254.  Piaget,  J.  The  origins  of  intelligence  in  children.  New  York:  Inter- 
national Universities  Press,  1952. 

255.  Piaget,  J.  The  construction  of  reality  in  the  child.  New  York:  Basic 
Books,  1954. 

256.  Piaget,  J.  Playy  dreams  and  imitation  in  childhood.  New  York: 
Norton,  1951. 

257.  Poetzl,  O.  Experimentell  erregte  Traumbilder  in  ihren  Beziehungen 
zum  indirekten  Sehen.  Z.  Neurol.  Psychiat.,  1917,  37,  278-349. 

258.  Rapaport,  D.  Emotions  and  memory.  (2nd  ed.)  New  York:  Inter- 
national Universities  Press,  1950. 

259.  Rapaport,  D.  Principles  underlying  projective  techniques.  Charact. 
&Pers.}  1942,  10,213-219. 

260.  Rapaport,  D.  Scientific  methodology  of  psychoanalysis.  Six  lectures 
given  at  the  Topeka  Institute  of  Psychoanalysis,  1945.  Unpublished  manu- 

261.  Rapaport,   D.    Principles  underlying  non-pro jective  tests   of  per- 
sonality. Ann.  N.  Y.  Acad.  Sd.,  1946, 46,  643-652. 

262.  Rapaport,    D.    Dynamic   psychology    and    Kantian    epistemology. 
Paper  presented  at  the  staff  seminar  of  the  Menninger  Foundation  School 
of  Clinical  Psychology,  1947.  Unpublished  manuscript. 


263.  Rapaport,   D.    Interpersonal   relations,   communication   and   psy- 
chodynamics.    Paper    presented    at    the    Menninger    Foundation    General 
Seminar,  1949.  Unpublished  manuscript. 

264.  Rapaport,   D.    Book   review:    "Cybernetics"    by   Norbert   Wiener. 
Psychoanal  Quart.,  1950,  19,  598-603. 

265.  Rapaport,  D.  On  the  psycho-analytic  theory  of  thinking.  Int.  J. 
Psycho-Anal,  1950,  31,  161-170.  Also  in  R.  P.  Knight  &  C.  R.  Friedman 
(Eds.) 5  Psychoanalytic  psychiatry  and  psychology,  clinical  and  theoretical 
papers.  Austen  Riggs  Center.  Vol.  L  New  York:  International  Universities 
Press,  1954.  Pp.  259-273. 

266.  Rapaport,  D.  The  autonomy  of  the  ego.  Bull.  Menninger  Clin., 
1951,  15,  113-123.  Also  in  R.  P.  Knight  &  C.  R.  Friedman  (Eds.),  Psycho- 
analytic psychiatry  and  psychology,  clinical  and  theoretical  papers.  Austen 
Riggs  Center.  Vol.   1.  New  York:   International  Universities  Press,   1954. 
Pp.  248-258. 

267.  Rapaport,  D.  The  conceptual  model  of  psychoanalysis.  /.  Pers., 
1951,  20,  56-81.  Also  in  R.  P.  Knight  &  C.  R.  Friedman  (Eds.),  Psycho- 
analytic psychiatry  and  psychology,  clinical  and  theoretical  papers.  Austen 
Riggs  Center.  Vol.   1.  New  York:    International  Universities  Press,   1954. 
Pp.  221-247. 

268.  Rapaport,  D.  (Ed.)   Organization  and  pathology  of  thought.  New 
York:  Columbia  Univer.  Press,  1951. 

269.  Rapaport,  D.  Paul  Schilder's  contribution  to  the  theory  of  thought- 
processes.   Int.   J.   Psycho-Anal,    1951,    32,    291-301.    Also,   Appendix    in 
Schilder,  P.,  Medical  psychology.  New  York:    International  Universities 
Press,  1953.  Pp.  340-356. 

270.  Rapaport,   D.   Consciousness:    a   psychopathological   and  psycho- 
dynamic  view.  In  Problems  of  consciousness.  Transactions  of  the  Second 
Conference,  March  19-20.  New  York:  Josiah  Macy,  Jr.  Foundation,  1951. 
Pp.  18-57. 

271.  Rapaport,  D.  Book  review:  O.  Hobart  Mowrer,  "Learning  theory 
and  personality  dynamics."  /.  abnorm.  soc.  PsychoL,  1952,  47,  137-142. 

272.  Rapaport,  D.  Book  review:    J.  Bollard  and  N.  E.  Miller,    "Per- 
sonality and  psychotherapy:  an  analysis  in  terms  of  learning,  thinking,  and 
culture."  Amer.  J.  Orthopsychiat.,  1953,  23,  204-208. 

273.  Rapaport,  D.  Discussion  in  H.  Powdermaker   (Ed.),  Mass  com- 
munications seminar:    proceedings   of   an    interdisciplinary   seminar.    New 
York:  Wermer-Gren  Foundation,  1953.  Pp.  121-128. 

274.  Rapaport,   D.   On   the   psycho-analytic  theory  of  affects.   Int.  J. 
Psycho-Anal,  1953,  34,  177-198.  Also  in  R.  P.  Knight  &  C.  R.  Friedman 
(Eds.),  Psychoanalytic  psychiatry  and  psychology,  clinical  and  theoretical 
papers.  Austen  Riggs  Center.  Vol.  1.  New  York:  International  Universities 
Press,  1954.  Pp.  274-310. 

275.  Rapaport,  D.   Some  metapsychological   considerations  concerning 
activity  and  passivity.  Two  lectures  given  at  the  staff  seminar  of  the  Austen 
Riggs  Center,  1953.  Unpublished  manuscript. 

The  Structure  of  Psychoanalytic  Theory  181 

276.  Rapaport  D.  Cognitive  structures.  In  Contemporary  approaches  to 
cognition:   a  symposium   held  at  the   University  of  Colorado.   Cambridge, 
Mass.:  Harvard  Univer.  Press,  1957. 

277.  Rapaport  D.  The  development  and  the  concepts  of  psychoanalytic 
ego-psychology.  Twelve  seminars  given  at  the  Western  New  England  Insti- 
tute for  Psychoanalysis,  1955.  Unpublished  manuscript  (mimeo.'l . 

278.  Rapaport,  D.  An  historical  survey  of  psychoanalytic  ego  psychology. 
Bull.  Philadelphia  Ass.  Psychoanal,  1958?  8,  105-120. 

279.  Rapaport,  D.  Psychoanalysis  and  developmental  psychology.  Lecture 
given  at  Clark  Univer.,  Worcester,  Mass.,  Sept.  21,  1957. 

280.  Rapaport,  D.  The  theory  of  ego  autonomy:  a  generalization.  Bull. 
Menninger  Clin.,  1958,  22,  13-35. 

281.  Rapaport,  D.,  Gill,  M.  M.3  &  Schafer,  R.  Diagnostic  psychological 
testing.  Vols.  1  &  2.  Chicago:  Yearbook  Publishers,  1945,  1946. 

282.  Red!,  F.,  &  Wineman,  D.  Children  who  hate.  Glencoe,  III.:  Free 
Press,  1951.  Reprinted  with  [283]  in  The  treatment  of  childhood  aggression. 
New  York:  Basic  Books,  1957. 

283.  Redl,   F.,  &  Wineman,  D.   Controls  from  within.  Glencoe,  111.: 
Free   Press,    1952.  Reprinted  with   [282]   in   The   treatment  of  childhood 
aggression.  New  York:  Basic  Books,  1957. 

284.  Reich,  W.  Char akter analyse.  Vienna:   SelbstverJag  des  Verfassers, 

285.  Reider,  N.  A  type  of  psychotherapy  based  on  psychoanalytic  prin- 
ciples. Bull.  Menninger  Clin.,  1955,  19,  111-128. 

286.  Reik,  T.  From  thirty  years  with  Freud.  New  York:  International 
Universities  Press,  1940. 

287.  Rich,  J.  L.  Goals  and  trends  of  research  in  geology  and  geography. 
Science,  1948,  107,  581-584. 

288.  Roffenstein,  G.  Experiments  on  symbolization  in  dreams.  In  [268, 
pp.  249-256]. 

289.  Roheim,    G.    The    origin    and   function    of   culture.   New   York: 
Nervous  and  Mental  Disease  Monograph,  1943. 

290.  Roheim,  G.  (Ed.)  Psychoanalysis  and  the  social  sciences.  Vols.  1-4. 
New  York:  International  Universities  Press,  1947-1955. 

291.  Roheim,  G.  Psychoanalysis  and  anthropology.  New  York:    Inter- 
national Universities  Press,  1950. 

292.  Rorschach,  H.  Psycho  diagnostics:  a  diagnostic  test  based  on  per- 
ception. New  York:  Grune  &  Stratton,  1942. 

293.  Rosenzweig,  S.  The  experimental  study  of  psychoanalytic  concepts. 
Charact.  &  Pers.3  1937,  6,  61-71. 

294.  Rosenzweig,  S.  The  experimental  study  of  repression.   In  H.  A. 
Murray    (Ed.),  Explorations  in  personality.  New  York:    Oxford  Univer. 
Press,  1938.  Pp.  472-490. 

295.  Sanford,  R.  N.  The  effects  of  abstinence  from  food  upon  imaginal 
processes:   a  preliminary  experiment  /.  PsychoL,  1936,  2,  129-136. 

296.  Sanford,  R.  N.  The  effects  of  abstinence  from  food  upon  imaginal 
processes:  a  further  experiment.  /.  PsychoL,  1937,  3,  145-159. 


297.  de  Saussure,  R.  Ueber  genetische  Psychoiogie  und  Psychoanalyse. 
Imago,  1934,  20,  282-315. 

298.  Schafer,   R.   Psychoanalytic   interpretation   in   Rorschach    testing: 
theory  and  application.  New  York:  Grune  &  Stratton,  1954. 

299.  Schilder,  P.  Mind:  perception  and  thought  in  their  constructive 
aspects.  New  York:  Columbia  Univer.  Press,  1942. 

300.  Schilder,  P.  On  the  development  of  thoughts.  In  [268,  pp.  497- 

301.  Schilder,  P.  Studies  concerning  the  psychology  and  symptomatology 
of  general  paresis.  In  [268,  pp.  519-580]. 

302.  Schilder,  P.  Medical  psychology.  New  York:    International  Uni- 
versities Press,  1953. 

303.  Schiller,  C.  H.  (Ed.  &  Trans.)  Instinctive  behavior:  the  develop- 
ment of  a  modern  concept.  New  York:   International  Universities  Press, 

304.  Schlesinger,  H.  J.  Cognitive  attitudes  in  relation  to  susceptibility 
to  interference./.  Pers.,  1954,  22,  354-374. 

305.  Schroedinger,  E.  What  is  life?  New  York:  Macmillan,  1947. 

306.  Schroetter,  K.  Experimental  Dreams.  In  [268,  pp.  234-248]. 

307.  Sears,   R.   R.    Functional   abnormalities   of  memory  with   special 
reference  to  amnesia.  Psychol.  Bull,  1936,  33,  229-274. 

308.  Sears,  R.  R.  Survey  of  objective  studies  of  psychoanalytic  concepts. 
New  York:  Soc.  Sci.  Res.  Council,  Bulletin  No.  51,  1943. 

309.  Shakow,  D.,  &  Rapaport,  D.  The  influence  of  Freud  on  psychology. 
New  York :  International  Universities  Press,  in  press. 

310.  Silberer,  H.  Report  on  a  method  of  eliciting  and  observing  certain 
symbolic  hallucination-phenomena.  In  [268,  pp.  195-207]. 

311.  Simpson,,  G.  G.  Quantitative  zoology:  numerical  concepts  in  the 
study  of  recent  and  fossil  animals.   New  York:    McGraw-Hill,    1939. 

312.  Simpson,  G.  G.  Tempo  and  mode  in  evolution.  New  York:  Colum- 
bia Univer.  Press,  1949. 

313.  Sullivan,   H.   S.   Conceptions  of  modern  psychiatry.  Washington, 
B.C.:  William  Alanson  White  Psychiatric  Foundation,  1947. 

314.  Sullivan,  H.  S.  The  interpersonal  theory  of  psychiatry.  New  York: 
Norton,  1953. 

315.  Tinbergen,  H.   The  study  of  instinct.  Oxford:    Clarendon  Press, 

316.  Tolman,  E.  C.  The  intervening  variable.  In  M.  Marx  (Ed.),  Psy- 
chological theory.  New  York:  Macmillan,  1951.  Pp.  87-102. 

317.  Travell,  J.,  &  Bigelow,  N.  H.  Role  of  somatic  trigger  areas  in 
the  patterns  of  hysteria.  Psychosom.  Med.3  1947,  9,  353-363. 

318.  Troland,  L.  T.  The  fundamentals  of  human  motivation.  Princeton, 
N.J.:  VanNostrand,  1928. 

319.  Varendonck,  J.  The  psychology  of  daydreams.  New  York:   Mac- 
millan, 1921.  Also  (abridged)  in  [268,  pp.  451-473]. 

320.  Waelder,  R.  The  principle  of  multiple  function:   observations  on 
over-determination.  PsychoanaL  Quart.,,  1936,  5,  45-62. 

Tht   Struct u re  of  Psychoanalytic  Thtury  183 

321.  Weiss,   E..  &  English,  O.  S.  P-^cJ-^o-^stic  rr^ditiKf:   ttic  clinical 

application  of  psycho  pat  fio!:gy  ij  g^r.crcl  mfdiccl  problem  \  Philadelphia: 
Saunders,  1943. 

322.  Werner.  H.   Comparative  psychology  of  mental  development.    (3d 
ed.)  New  York:  Internationa!  Universities  Press..  1957. 

323.  Werner,  BL  &  Wapner,  S.  Toward  a  general  theory  of  perception. 
PsychoL  Rev.,  1952,  593  324-338. 

324.  Wexler,  M.  The  structural  problem  in  schizophrenia:   therapeutic 
implications.  Int.  /.  Psycho-Anal.,  1951,  325  157-166. 

325.  Wexler,  M.  The  structural  problem  in  schizophrenia:   the  role  of 
the  internal  object.  Bull.  Menninger  Clin.3  1951,  15,  221-234. 

326.  Wiener,  N.  Cybernetics.  New  York:  Wiley.  1948. 

327.  Wittels,  F.  Sigmund  Freud:  his  personality y  his  teaching  and  his 
school  New  York:  Dodd,  Mead,  1924. 

328.  Wolff,   P.   H.   Piaget's  genetic  psychology  and  its  relation  to  psy- 
choanalysis. Unpublished  manuscript. 

329.  Woodworth,   R.    S.    Dynamic   psychology.    New   York:    Columbia 
Univer.  Press,  1925. 

330.  ZetzeL  E.  The  depressive  position.  In  P.  Greenacre  (Ed.),  Affec- 
tive disorders.  New  York:  International  Universities  Press,  1953.  Pp.  84—116. 



University  of  Wisconsin 

Introduction  {1} 185 

The  soil  of  the  theory 185 

Some  basic  attitudes 188 

The  General  Structure  of  Our  Systematic  Thinking  {2  +,  3} 192 

Definitions  of  constructs 194 

A  digression  on  the  case  history  of  a  construct  {3-f  }• 200 

I.  A  Theory  of  Therapy  and  Personality  Change  {2+,  6,  8,  9}    .      .      .      .  212 

Conditions  of  the  therapeutic  process 213 

The  process  of  therapy 216 

Outcomes  in  personality  and  behavior 2'18 

Comments  on  the  theory  of  therapy 220 

Specification  of  functional  relationships 220 

Some  conclusions  regarding  the  nature  of  the  individual 220 

II.  A  Theory  of  Personality  {2  +,  6,  9} 221- 

Postulated  characteristics  of  the  human  infant 222 

The  development  of  the  self 223 

The  need  for  positive  regard 223 

The  development  of  the  need  for  self-regard 224 

The  development  of  conditions  of  worth 224 

The  development  of  incongruence  between  self  and  experience      .      .      .  226 

The  development  of  discrepancies  in  behavior 227 

The  experience  of  threat  and  the  process  of  defense 227 

The  process  of  breakdown  and  disorganization 228 

The  process  of  reintegration 230 

Specification  of  functional  relationships  in  the  theory  of  personality    .      .  231 

Evidence 232- 

III.  A  Theory  of  the  Fully  Functioning  Person  {2  +5  6} 234 


Therapy,  Personality,  and  Interpersonal  Relationships  185 

IV.  A  Theory  of  Interpersonal  Relationship  {2-r,  6} 235 

V.  Theories  of  Application  {2 -r,  6} 241 

Family  life 241 

Education  and  learning 241 

Group  leadership 24^ 

Group  tension  and  conflict 242 

The  Theoretical  System  in  a  Context  of  Research 244 

The  bases  of  stimulation  of  research  {8} 245 

The  problem  of  measurement  and  quantification  {5  > 246 

Incompatible  evidence  {9,  7} 247 

A  continuing  program  of  theory  and  research  {11,7} 249 

Immediate  strategy  of  development  {12} 250 

Conclusion 252 

References 252 


Being  one  who  has  deprecated  the  use  of  compulsion  as  a  means  of 
altering  personality  and  behavior,  it  is  no  doubt  singularly  appropriate 
that  I  should  be  forced  to  acknowledge  the  value  of  the  gentle  com- 
pulsion of  a  formal  request.  For  some  time  I  had  recognized  the  need 
of  a  more  adequate  and  more  up-to-date  statement  of  the  theories  which 
have  been  developing  in  the  group  associated  with  client-centered 
therapy.  This  might  well  have  remained  in  the  realm  of  good  intentions, 
had  it  not  been  for  the  formal  request  from  the  American  Psychological 
Association,  in  connection  with  its  Study  of  the  Status  and  Development 
of  Psychology  in  the  United  States,  to  prepare  a  systematic  statement 
of  this  developing  theory.  To  join  with  others  who  were  endeavoring  to 
formulate  their  own  theories  and  to  use,  so  far  as  possible,  a  common 
outline — this  seemed  to  be  both  an  obligation  and  an  opportunity  which 
could  not  be  refused.  It  is  this  softly  voiced  but  insistent  pressure  from 
my  colleagues  which  has  caused  me  to  write  the  following  pages  now, 
rather  than  at  some  later  date.  For  this  pressure  I  am  grateful. 

The  soil  of  the  theory.  No  theory  can  be  adequately  understood 
without  some  knowledge  of  the  cultural  and  personal  soil  from  which 
it  springs.  Consequently  I  am  pleased  that  the  first  item  of  the  suggested 
outline  requests  a  thorough  discussion  of  background  factors.  This  means, 
I  fear,  that  I  must  take  the  reader  through  some  autobiographical  ma- 
terial since,  although  the  client-centered  orientation  has  become  very 
much  of  a  group  enterprise  in  every  respect,  I,  as  an  individual,  carry  a 
considerable  responsibility  for  its  initiation  and  for  the  beginning 
formulation  of  its  theories.  I  shall,  therefore,  mention  briefly  some 
cultural  influences  and  personal  experiences  which  may  or  may  not 



have  relevance  to  the  theory  itself.  I  shall  not  attempt  to  evaluate  these 
influences,  since  1  am  probably  a  poor  judge  of  the  part  they  have 

I  lived  my  childhood  as  a  middle  child  in  a  large,  close-knit  family, 
where  hard  work  and  a  highly  conservative  (almost  fundamentalist) 
Protestant  Christianity  were  about  equally  revered.  When  the  family 
moved  to  a  farm  at  the  time  I  was  twelve,  I  became  deeply  interested 
and  involved  in  scientific  agriculture.  The  heavy  research  volumes  I  read 
on  my  own  initiative  in  the  next  few  years  regarding  feeds  and  feeding, 
soils,  animal  husbandry,  and  the  like,  instilled  in  me  a  deep  and  abiding 
respect  for  the  scientific  method  as  a  means  of  solving  problems  and 
creating  new  advances  in  knowledge.  This  respect  was  reinforced  by  rny 
first  years  in  college,  where  I  was  fond  of  the  physical  and  biological 
sciences.  In  my  work  in  history  I  also  realized  something  of  the  satis- 
factions of  scholarly  work. 

Having  rejected  the  family  views  of  religion,  I  became  interested  in 
a  more  modern  religious  viewpoint  and  spent  two  profitable  years  in 
Union  Theological  Seminary,  which  at  that  time  was  deeply  committed 
to  a  freedom  of  philosophical  thought  which  respected  any  honest  at- 
tempt to  resolve  significant  problems,  whether  this  led  into  or  away  from 
the  church.  My  own  thinking  lead  me  in  the  latter  direction,  and  I 
moved  "across  the  street"  to  Teachers  College,  Columbia  University. 
Here  I  was  exposed  to  the  views  of  John  Dewey,  not  directly,  but 
through  William  H.  KilpatricL  I  also  had  my  first  introduction  to 
clinical  psychology  in  the  warmly  human  and  common-sense  approach 
of  Leta  Hollingworth.  There  followed  a  year  of  internship  at  the  In- 
stitute for  Child  Guidance,  then  in  its  chaotic  but  dynamic  first  year  of 
existence.  Here  I  gained  much  from  the  highly  Freudian  orientation  of 
most  of  its  psychiatric  staff,  which  included  David  Levy  and  Lawson 
Lowrey.  My  first  attempts  at  therapy  were  carried  on  at  the  Institute. 
Because  I  was  still  completing  my  doctorate  at  Teachers  College,  the 
sharp  incompatibility  of  the  highly  speculative  Freudian  thinking  of  the 
Institute  with  the  highly  statistical  and  Thorndikean  views  at  Teachers 
College  was  keenly  felt. 

There  followed  twelve  years  in  what  was  essentially  a  community 
child  guidance  clinic  in  Rochester,  New  York.  This  was  a  period  of 
comparative  isolation  from  the  thinking  of  others.  The  psychology  de- 
partment of  the  University  of  Rochester  was  uninterested  in  what  we 
were  doing  because  our  work  was  not,  in  its  opinion,  in  the  field  of 
psychology.  Our  colleagues  in  the  social  agencies,  schools,  and  courts 
knew  little  and  cared  less  about  psychological  ideologies.  The  only 
element  which  carried  weight  with  them  was  the  ability  to  get  results 
in  working  with  maladjusted  individuals.  The  staff  was  eclectic,  of  diverse 

Therapy,  Personality,  and  Interpersonal  Relationships  187 

background,  and  our  frequent  and  continuing  discussion  of  treatment 
methods  was  based  on  our  practical  even-day  working  experience  with 
the  children,  adolescents,  and  adults  who  were  our  clients.  It  was  the 
beginning  of  an  effort,  which  has  had  meaning  for  rue  ever  since,  to 
discover  the  order  which  exists  in  our  experience  of  working  with  people. 
The  volume  on  the  Clinical  Treatment  of  the  Problem  Child  was  one 
outcome  of  this  effort. 

During  the  second  half  of  this  period  there  were  several  individuals 
who  brought  Into  our  group  the  controversial  therapeutic  views  of  Otto 
Rank  and  the  Philadelphia  group  of  social  workers  and  psychiatrists 
whom  he  had  Influenced.  Personal  contact  with  Rank  was  limited  to 
a  three-day  institute  we  arranged;  nevertheless  his  thinking  had  a  very 
decided  impact  on  our  staff  and  helped  me  to  crystallize  some  of  the 
therapeutic  methods  we  were  groping  toward.  For  by  this  time  I  was 
becoming  more  competent  as  a  therapist,  and  beginning  to  sense  a  dis- 
coverable orderliness  in  this  experience,  an  orderliness  which  was  in- 
herent in  the  experience,  and  (unlike  some  of  the  Freudian  theories 
which  had  grown  so  far  from  their  original  soil)  did  not  have  to  be 
imposed  on  the  experience. 

Though  I  had  earned  on  some  part-time  university  teaching  through- 
out the  Rochester  years,  the  shift  to  a  faculty  position  at  Ohio  State 
University  w^as  a  sharp  one.  I  found  that  the  emerging  principles  of 
therapy,  which  I  had  experienced  largely  on  an  implicit  basis,  were  by 
no  means  clear  to  well-trained,  critically  minded  graduate  students.  I 
began  to  sense  that  what  I  was  doing  and  thinking  In  the  clinical  field 
was  perhaps  more  of  a  new  pathway  than  I  had  recognized.  The  paper 
I  presented  to  the  Minnesota  chapter  of  Psi  Chi  in  December,  1940, 
(later  chapter  2  of  Counseling  and  Psychotherapy)  was  the  first  con- 
scious attempt  to  develop  a  relatively  new  line  of  thought.  Up  to  that 
time  I  had  felt  that  my  writings  were  essentially  attempts  to  distill  out 
more  clearly  the  principles  which  "all  clinicians"  were  using. 

The  new  influence  at  Ohio  State,  which  continued  to  be  felt  in  my 
years  at  Chicago,  was  the  impact  of  young  men  and  women — intellectu- 
ally curious,  often  theoretically  oriented,  eager  to  learn  from  experience 
and  to  contribute  through  research  and  theory  to  the  development  of  a 
field  of  knowledge.  Through  their  mistakes  as  well  as  their  successes  in 
therapy,  through  their  research  studies,  their  critical  contributions,  and 
through  our  shared  thinking,  have  come  many  of  the  recent  develop- 
ments in  this  orientation. 

In  the  past  decade  at  the  University  of  Chicago  the  new  elements 
which  stand  out  most  sharply  are  the  opportunity  for  and  the  encourage- 
ment of  research,  the  inclusion  of  graduate  students  from  education, 
theology,  human  development,  sociology,  industrial  relations,  as  well  as 

188  CARL   R.    ROGERS 

psychology,  in  the  ramified  activities  of  the  Counseling  Center,  and  the 
creative  thinking  of  my  faculty  colleagues,  especially  those  connected 
with  the  Center. 

The  persistent  influence  which  might  not  be  fully  recognized,  because 
it  is  largely  implicit  in  the  preceding  paragraphs,  is  the  continuing 
clinical  experience  with  individuals  who  perceive  themselves,  or  are 
perceived  by  others  to  be,  in  need  of  personal  help.  Since  1928,  for  a 
period  now  approaching  thirty  years,  I  have  spent  probably  an  average 
of  15  to  20  hr  per  week,  except  during  vacation  periods,  in  endeavoring 
to  understand  and  be  of  therapeutic  help  to  these  individuals.  To  me, 
they  seem  to  be  the  major  stimulus  to  my  psychological  thinking.  From 
these  hours,  and  from  my  relationships  with  these  people,  I  have  drawn 
most  of  whatever  insight  I  possess  into  the  meaning  of  therapy,  the 
dynamics  of  interpersonal  relationships,  and  the  structure  and  function- 
ing of  personality. 

Some  basic  attitudes.  Out  of  this  cultural  and  personal  soil  have 
grown  certain  basic  convictions  and  attitudes  which  have  undoubtedly 
influenced  the  theoretical  formulation  which  will  be  presented.  I  will 
endeavor  to  list  some  of  these  views  which  seem  to  me  relevant : 

1.  I  have  come  to  see  both  research  and  theory  as  being  aimed 
toward  the  inward  ordering  of  significant  experience.  Thus  research 
is  not  something  esoteric,  nor  an  activity  in  which  one  engages  to  gain 
professional  kudos.  It  is  the  persistent,  disciplined  effort  to  make  sense 
and  order  out  of  the  phenomena  of  subjective  experience.  Such  effort  is 
justified  because  it  is  satisfying  to  perceive  the  world  as  having  order  and 
because  rewarding  results  often  ensue  when  one  understands  the  orderly 
relationships  which  appear  to  exist  in  nature.  One  of  these  rewarding 
results  is  that  the  ordering  of  one  segment  of  experience  in  a  theory  im- 
mediately opens  up  new  vistas  of  inquiry,  research,  and  thought,  thus 
leading  one  continually  forward. 

Thus  the  primary  reason  for  research  and  systematic  theory  in  the 
field  of  therapy  is  that  it  is  personally  dissatisfying  to  permit  the  cumulat- 
ing experiences  of  therapeutic  hours  to  remain  as  a  conglomeration  of 
more  or  less  isolated  events.  It  feels  as  though  there  is  an  order  in  these 
events.  What  could  it  be?  And  of  any  hunch  regarding  the  inherent 
order,  it  is  necessary  to  ask  the  question,  is  this  really  true,  or  am  I 
deceiving  myself?  Thus  slowly  there  is  assembled  a  body  of  facts,  and 
systematic  constructs  to  explain  those  facts,  which  have  as  their  basic 
function  the  satisfaction  of  a  need  for  order  which  exists  in  me. 

(I  have,  at  times,  carried  on  research  for  purposes  other  than  the 
above  to  satisfy  others,  to  convince  opponents  and  sceptics,  to  gain 
prestige,  and  for  other  unsavory  reasons.  These  errors  in  judgment  and 
activity  have  only  deepened  the  above  positive  conviction. ) 

Therapy,  Personality,  and  Interpersonal  Relationships  189 

2.  It  is  my  opinion  that  the  type  of  understanding  which  we  call 
science  can  begin  anywhere,  at  any  level  of  sophistication.  To  observe 
acutely,    to    think   carefully   and   creatively — these    activities,    not   the 
accumulation  of  laboratory  instruments,  are  the  beginnings  of  science. 
To  observe  that  a  given  crop  grows  better  on  the  rocky  hill  than  in  the 
lush  bottom  land,  and  to  think  about  this  observation,  is  the  start  of 
science.  To  notice  that  most  sailors  get  scurvy  but  not  those  who  have 
stopped  at  islands  to  pick  up  fresh  fruit  is  a  similar  start.  To  recognize 
that,  when  a  person's  views  of  himself  change,  his  behavior  changes 
accordingly,   and  to  puzzle  over  this,  is  again  the  beginning  of  both 
theory7  and  science.  I  voice  this  conviction  in  protest  against  the  attitude, 
which  seems  too  common  in  American  psychology,  that  science  starts  in 
the  laboratory  or  at  the  calculating  machine. 

3.  A  closely  related  belief  is  that  there  is  a  natural  history  of  science 
— that  science,  in  any  given  field,  goes  through  a  patterned  course  of 
growth  and  development.  For  example,  it  seems  to  me  right  and  natural 
that  in  any  new  field  of  scientific  endeavor  the  observations  are  gross, 
the  hypotheses  speculative  and  full  of  errors,  the  measurements  crude. 
More  important,  I  hold  the  opinion  that  this  is  just  as  truly  science  as  the 
use  of  the  most  refined  hypotheses  and  measurements  in  a  more  fully 
developed  field  of  study.  The  crucial  question  in  either  case  is  not  the 
degree  of  refinement  but  the  direction  of  movement.  If  in  either  instance 
the  movement  is  toward  more  exact  measurement,  toward  more  clear- 
cut  and  rigorous  theory  and  hypotheses,  toward  findings  which  have 
greater  validity  and  generality,  then  this  is  a  healthy  and  growing  science. 
If  not,  then  it  is  a  sterile  pseudo  science,  no  matter  how  exact  its 
methods.  Science  is  a  developing  mode  of  inquiry,  or  it  is  of  no  par- 
ticular importance. 

4.  In  the  invitation  to  participate  in  the  APA  study,  I  have  been 
asked  to  cast  our  theoretical  thinking  in  the  terminology  of  the  in- 
dependent-intervening-dependent  variable,  in  so  far  as  this  is  feasible. 
I  regret  that  I  find  this  terminology  somehow  uncongenial.  I  cannot 
justify  my  negative  reaction  very  adequately,   and  perhaps  it  is  an 
irrational  one,  for  the  logic  behind  these  terms  seems  unassailable.  But 
to  me  the  terms  seem  static — they  seem  to  deny  the  restless,  dynamic, 
searching,  changing  aspects  of  scientific  movement.  There  is  a  tendency 
to  suppose  that  a  variable  thus  labeled,  remains  so,  which  is  certainly 
not  true.  The  terms  also  seem  to  me  to  smack  too  much  of  the  laboratory, 
where  one  undertakes  an  experiment  de  novo,  with  everything  under 
control,  rather  than  of  a  science  which  is  endeavoring  to  wrest  from 
the  phenomena  of  experience  the  inherent  order  which  they  contain. 
Such  terms  seem  to  be  more  applicable  to  the   advanced  stages  of 
scientific  endeavor  than  to  the  beginning  stages. 

190  CARL    R.    ROGERS 

Please  do  not  misunderstand.  I  quite  realize  that  after  the  fact,  any 
research  investigation,  or  any  theory  constructed  to  relate  the  discovered 
facts,  should  be  translatable  into  the  language  of  independent  and 
dependent  variables  or  there  is  something  wrong  with  the  research  or 
theory.  But  the  terms  seem  to  me  better  adapted  to  such  autopsies  than 
to  the  living  physiology  of  scientific  work  in  a  new  field, 

5.  It  should  be  quite  clear  from  the  foregoing  that  the  model  of 
science  which  I  find  most  helpful  is  not  taken  from  the  advanced  stages 
of  theoretical  physics.  In  a  field  such  as  psychotherapy  or  personality 
the  model  wiiich  seems  more  congenial  to  me  would  be  taken  from  the 
much  earlier  stages  of  the  physical  sciences.  I  like  to  think  of  the  dis- 
covery of  radioactivity  by  the  Curies.  They  had  left  some  pitchblende 
ore,  which  they  were  using  for  some  purpose  or  other,  in  a  room  where 
they  stored  photographic  plates.  They  discovered  that  the  plates  had 
been  spoiled.    In   other  words,   first  there  was  the  observation   of   a 
dynamic  event.  This  event  might  have  been  due  to  a  multitude  of 
causes.  It  might  have  been  a  flaw  in  the  manufacture  of  the  plates.  It 
might  have  been  the  humidity,  the  temperature,  or  any  one  of  a  dozen 
other  things.  But  acute  observation  and  creative  thinking  fastened  on 
a  hunch  regarding  the  pitchblende,  and  this  became  a  tentative  hypoth- 
esis. Crude  experiments  began  to  confirm  the  hypothesis.  Only  slowly 
was  it  discovered  that  it  was  not  the  pitchblende,  but  a  strange  element 
in  the  pitchblende  which  was  related  to  the  observed  effect.  Meanwhile 
a  theory  had  to  be  constructed  to  bring  this  strange  phenomenon  into 
orderly  relationship  with  other  knowledge.  And  although  the  theory  in  its 
most  modest  form  had  to  do  with  the  effect  of  radium  on  photographic 
plates,  in  its  wider  and  more  speculative  reaches  it  was  concerned  with 
the  nature  of  matter  and  the  composition  of  the  universe.  By  present-day 
standards  in  the  physical  sciences,  this  is  an  example  of  a  primitive  stage 
of  investigation  and  theory  construction.  But  in  the  fields  in  which  I  am 
most  deeply  interested  I  can  only  hope  that  we  are  approaching  such  a 
stage.  I  feel  sure  that  we  are  not  beyond  it. 

6.  Another  deep-seated  opinion  has  to  do  with  theory.  I  believe  that 
there  is  only  one  statement  which  can  accurately  apply  to  all  theories — 
from  the  phlogiston  theory  to  the  theory  of  relativity,  from  the  theory 
I  will  present  to  the  one  which  I  hope  will  replace  it  in  a  decade — and 
that  is  that  at  the  time  of  its  formulation  every  theory  contains  an  un- 
known  (and  perhaps  at  that  point  an  unknowable)   amount  of  error 
and  mistaken  inference.  The  degree  of  error  may  be  very  great,  as  in 
the  phlogiston  theory,  or  small,  as  I  imagine  it  may  be  in  the  theory 
of  relativity,  but  unless  we  regard  the  discovery  of  truth  as  a  closed  and 
finished  book,  then  there  will  be  new  discoveries  which  will  contradict 
the  best  theories  which  we  can  now  construct. 

Therapy  y  Personality,  and  Interpersonal  Relationships  191 

To  me  this  attitude  Is  very  Important,  for  I  am  distressed  at  the 
manner  In  which  small-caliber  minds  Immediately  accept  a  theory — 
almost  any  theory — as  a  dogma  of  truth.  If  theory  could  be  seen  for  what 
it  Is — a  fallible,  changing  attempt  to  construct  a  network  of  gossamer 
threads  which  will  contain  the  solid  facts — then  a  theory  would  serve 
as  it  should,  as  a  stimulus  to  further  creative  thinking. 

I  am  sure  that  the  stress  I  place  on  this  grows  In  part  out  of  my 
regret  at  the  history  of  Freudian  theory.  For  Freud,  It  seems  quite  clear 
that  his  highly  creative  theories  were  never  more  than  that.  He  kept 
changing,  altering,  revising,  giving  new  meaning  to  old  terms — always 
with  more  respect  for  the  facts  he  observed  than  for  the  theories  he 
had  built.  But  at  the  hands  of  insecure  disciples  (so  It  seems  to  me), 
the  gossamer  threads  became  Iron  chains  of  dogma  from  which  dynamic 
psychology  Is  only  recently  beginning  to  free  Itself.  I  feel  that  every 
formulation  of  a  theory  contains  this  same  risk  and  that,  at  the  time 
a  theory  is  constructed,  some  precautions  should  be  taken  to  prevent  it 
from  becoming  dosma. 

o          o 

7.  I  share  with  many  others  the  belief  that  truth  Is  unitary,  even 
though  we  will  never  be  able  to  know  this  unity.  Hence  any  theory, 
derived  from  almost  any  segment  of  experience.  If  It  wrere  complete  and 
completely  accurate,  could  be  extended  Indefinitely  to  provide  meaning 
for  other  very  remote  areas  of  experience.  Tennyson  expressed  this  In 
sentimental  fashion  in  his  "Flower  In  the  Crannied  Wall."  I  too  believe 
that  a  complete  theory  of  the  Individual  plant  would  show  us  "what 
God  and  man  is." 

The  corollary,  however,  is  of  equal  importance  and  Is  not  so  often 
stated.  A  slight  error  in  a  theory  may  make  little  difference  in  providing 
an  explanation  of  the  observed  facts  out  of  which  the  theory  grew. 
But  wrhen  the  theory  Is  projected  to  explain  more  remote  phenomena, 
the  error  may  be  magnified,  and  the  inferences  from  the  theory  may  be 
completely  false.  A  very  slight  error  in  the  understanding  of  Tennyson's 
flower  may  give  a  grossly  false  understanding  of  man.  Thus  every  theory 
deserves  the  greatest  respect  In  the  area  from  which  it  was  drawn  from 
the  facts  and  a  decreasing  degree  of  respect  as  it  makes  predictions  in 
areas  more  and  more  remote  from  its  origin.  This  is  true  of  the  theories 
developed  by  our  owrn  group. 

8.  There  is  one  other  attitude  which  I  hold,  which  I  believe  has 
relevance  for  the  proper  evaluation  of  any  theory  I  might  present.  It  is 
my  belief  in  the  fundamental  predominance  of  the  subjective.  Man  lives 
essentially  in  his  own  personal  and  subjective  world,  and  even  his  most 
objective  functioning,  in  science,  mathematics,  and  the  like,  is  the  result 
of  subjective  purpose  and  subjective  choice.  In  relation  to  research  and 
theory,  for  example,  It  is  my  subjective  perception  that  the  machinery  of 

192  CARL    R.    ROGERS 

science  as  we  know  it — operational  definitions,  experimental  method, 
mathematical  proof — is  the  best  way  of  avoiding  self-deception.  But  I 
cannot  escape  the  fact  that  this  is  the  way  it  appears  to  me,  and  that  had 
I  lived  two  centuries  ago,  or  if  I  were  to  live  two  centuries  in  the 
future,  some  other  pathway  to  truth  might  seem  equally  or  more  valid. 
To  put  it  more  briefly,  it  appears  to  me  that  though  there  may  be  such 
a  thing  as  objective  truth,  I  can  never  know  it;  all  I  can  know  is  that 
some  statements  appear  to  me  subjectively  to  have  the  qualifications  of 
objective  truth.  Thus  there  is  no  such  thing  as  Scientific  Knowledge; 
there  are  only  individual  perceptions  of  what  appears  to  each  person 
to  be  such  knowledge. 

Since  this  is  a  large  and  philosophical  issue,  not  too  closely  related  to 
what  follows,  I  shall  not  endeavor  to  state  it  more  fully  here  but  refer 
any  who  are  interested  to  an  article  in  which  I  have  tried  to  expound 
this  view  somewhat  more  fully  [67].  I  mention  it  here  only  because 
it  is  a  part  of  the  context  in  which  my  theoretical  thinking  has  developed. 


Before  proceeding  to  the  detailed  statement  of  some  of  our  theoretical 
\iews,  I  believe  it  may  be  helpful  to  describe  some  of  the  interrelation- 
ships between  various  portions  of  our  theoretical  formulations. 

The  earliest  portion,  most  closely  related  to  observed  fact,  most 
heavily  supported  by  evidence,  is  the  theory  of  psychotherapy  and 
personality  change  which  was  constructed  to  give  order  to  the  phenomena 
of  therapy  as  we  experienced  it. 

In  this  theory  there  were  certain  hypotheses  regarding  the  nature 
of  personality  and  the  dynamics  of  behavior.  Some  of  these  were  ex- 
plicit, some  implicit.  These  have  been  developed  more  fully  into  a 
theory  of  personality.  The  purpose  has  been  to  provide  ourselves  with  a 
tentative  understanding  of  the  human  organism  and  its  developing 
dynamics — an  attempt  to  make  sense  of  this  person  who  comes  to  us  in 

Implicit  in  the  theories  of  therapy  and  of  personality  are  certain 
hypotheses  regarding  the  outcomes  of  therapy — hence,  hypotheses  re- 
garding a  more  socially  constructive  or  creative  individual.  In  the  last 
few  years  we  have  endeavored  to  spell  out  the  picture  of  the  theoretical 
end  point  of  therapy,  the  maximally  creative,  self-actualizing,  or  fully 
functioning  person. 

In  another  direction,  our  understanding  of  the  therapeutic  rela- 
tionship has  led  us  to  formulate  theoretical  statements  regarding  all 
interpersonal  relationships,  seeing  the  therapeutic  relationship  simply 

Therapy,  Personality,  and  Interpersonal  Relationships  193 

as  one  special  case.  This  is  a  very  new  and  tentative  development,  which 
we  believe  has  promise. 

Finally,  it  has  seemed  that  if  our  \iews  of  therapy  have  any  validity 
they  have  application  in  all  those  fields  of  human  experience  and  en- 
deavor which  involve  (a]  interpersonal  relationships  and  fb]  the  aim 
or  potentiality  of  development  or  change  in  personality  and  behavior. 

n.  A  THEORY 


H.A  1,2,3,4,5,6 
B  1,2 
C  1 

D  1,2,3,4 
E  1,2,3 
F  1,2,3 
G  1 

H  1,2,3,4 
I  1,2,3,4 
J  1,2,3, 


The  nature  of  the  human  organism 


3E                   3ZE                    "VTT  vi  n 



FlG.    I 

Consequently  a  cluster  of  partially  developed  theories  exists  in  relation 
to  such  fields  as  family  life,  education,  group  leadership,  and  situations 
of  group  tension  and  conflict. 

The  accompanying  chart  may  help  the  reader  to  see  and  understand 
these  relationships  between  different  aspects  of  our  theories.  It  should 
be  clear  that  the  chart  reads  from  the  center,  and  that  the  developments 
have  taken  place  in  the  four  directions  indicated.  It  should  also  be 
remembered  that  the  possibility  of  magnification  of  error  in  the  theory 
increases  as  one  goes  out  from  the  center.  By  and  large,  there  is  less 

194  CARL    R.    ROGERS 

evidence  available  in  these  peripheral  areas  than  in  the  center.  Entered 
in  the  chart  are  the  identifying  numbers  of  the  various  propositions 
which  follow,  so  that  in  reading  any  specific  portion  of  the  theory  the 
reader  may  refer  back  to  see  its  organic  relationship  to  other  parts  of 
the  theoretical  structure. 

Before  proceeding  to  set  forth  something  of  the  theories  themselves, 
I  should  like  gratefully  to  stress  the  extent  to  which  this  is  basically  a 
group  enterprise.  I  have  drawn  upon  specific  written  contributions  to 
theory  made  by  Victor  Raimy,  Richard  Hogan,  Stanley  Standal,  John 
Butler,  and  Thomas  Gordon.  Many  others  have  contributed  to  my 
thinking  in  ways  known  and  unknown,  but  I  would  particularly  like  to 
mention  the  valuable  influence  of  Oliver  Bown,  Desmond  Cartwright, 
Arthur  Combs,  Eugene  Gendlin,  A.  H.  Maslow,  Julius  Seeman,  John 
Shlien,  and  Donald  Snygg  on  the  theories  which  I  am  about  to  present. 
Yet  these  individuals  are  by  no  means  to  be  held  responsible  for  what 
follows,  for  their  own  attempts  to  order  experience  have  often  led  them 
into  somewhat  different  channels  of  thinking. 

Definitions  of  constructs.  In  the  development  of  our  theories  various 
systematic  constructs  have  emerged,  gradually  acquiring  sharper  and 
more  specific  meaning.  Also  terms  in  common  usage  have  gradually 
acquired  somewhat  specialized  meanings  in  our  theoretical  statements. 
In  this  section  I  have  endeavored  to  define,  as  rigorously  as  I  am  able, 
these  constructs  and  terms.  These  definitions  supply  the  means  by  which 
the  theory  may  be  more  accurately  understood. 

In  this  section  one  will  find  first  a  numbered  list  of  all  of  the  con- 
structs defined,  grouped  in  related  clusters.  There  are  eleven  of  these 
clusters,  each  with  a  focal  concept.  If  these  focal  concepts  are  under- 
stood, the  understanding  of  each  of  the  related  terms  should  not  be 
difficult,  since  each  of  the  constructs  within  a  group  has  a  close  and 
meaningful  relationship  to  the  others. 

Following  the  list  one  will  find  each  of  the  constructs  in  the  order 
numbered.  Each  is  defined,  and  explanatory  comment  is  often  added. 

In  connection  with  one  cluster  of  concepts,  those  having  to  do  with 
the  self,  there  is  a  long  digression  giving  the  "case  history"  of  the  develop- 
ment of  that  construct.  This  is  intended  to  illustrate  the  way  in  which 
most  of  the  constructs  in  this  theoretical  system  have  been  developed,  not 
as  armchair  constructs  but  out  of  a  continuing  interplay  between  thera- 
peutic experience,  abstract  conceptualizing,  and  research  using  opera- 
tionally defined  terms. 

It  is  quite  possible  that  such  a  section,  devoted  entirely  to  definitions, 
will  prove  dull  reading.  The  reader  may  prefer  to  go  at  once  to  the 
theory  of  therapy  in  the  following  section,  where  he  will  find  each  defined 
term  printed  in  italics.  He  may  then  refer  back  to  this  section  for  the 
exact  meaning  of  each  such  term. 

Therapy,  Personality,  and  Interpersonal  Relationships  195 
Grouping  of  Definitions 

Actualizing  tendency  and  related  construct 

1.  Actualizing  tendency 

2.  Tendency  toward  self-actualization 
Experience  and  related  constructs 

3.  Experience  f/noun) 

4.  Experience  (verb) 

5.  Feeling,  Experiencing  a  feeling 
Awareness  and  related  constructs 

6.  Awareness,  Symbolization,  Consciousness 

7.  Availability  to  awareness 

8.  Accurate  symbolization 

9.  Perceive,  Perception 

10.  Subceive,  Subception 
Self  and  related  constructs 

11.  Self-experience 

12.  Self,  Concept  of  self,  Self-structure 

13.  Ideal  self 
Incongruence  and  related  constructs 

14.  Incongruence  between  self  and  experience 

15.  Vulnerability 

16.  Anxiety 

17.  Threat 

18.  Psychological  maladjustment 
The  response  to  threat 

19.  Defense,  Defensiveness 

20.  Distortion  in  awareness,  Denial  to  awareness 

21.  Intensionality 
Congruence  and  related  constructs 

22.  Congruence  of  self  and  experience 

23.  Openness  to  experience 

24.  Psychological  adjustment 

25.  Extensionality 

26.  Mature,  Maturity 

Unconditional  positive  regard  and  related  constructs 

27.  Contact 

28.  Positive  regard 

29.  Need  for  positive  regard 

30.  Unconditional  positive  regard 
-  31.  Regard  complex 

32.  Positive  self-regard 

33.  Need  for  self-regard 

34.  Unconditional  self-regard 

196  CARL   R.    ROGERS 

Conditions  of  worth 

35.  Conditions  of  worth 
Constructs  related  to  valuing 

36.  Locus  of  evaluation 

37.  Organismic  valuing  process 
Constructs  related  to  source  of  knowledge 

38.  Internal  frame  of  reference 

39.  Empathy 

40.  External  frame  of  reference 

1.  Actualizing  tendency.  This  is  the  inherent  tendency  of  the  organ- 
ism to  develop  all  its  capacities  in  ways  which  serve  to  maintain  or  en- 
hance the  organism.  It  involves  not  only  the  tendency  to  meet  what 
Maslow  [45]  terms  "deficiency  needs"  for  air,  food,  water,  and  the  like, 
but  also  more  generalized  activities/It  involves  development  toward  the 
differentiation  of  organs  and  of  functions,  expansion  in  terms  of  growth, 
expansion  of  effectiveness  through  the  use  of  tools,  expansion,  and  en- 
hancement through  reproduction.  It  is  development  toward  autonomy 
and  away  from  heteronomy,  or  control  by  external  forces.  Angyal's  state- 
ment [2]  could  be  used  as  a  synonym  for  this  term:  "Life  is  an  auton- 
omous  event   which  takes   place   between   the   organism   and   the   en- 
vironment. Life  processes  do  not  merely  tend  to  preserve  life  but  tran- 
scend the  momentary  status  quo  of  the  organism,  expanding  itself  con- 
tinually and  imposing  its  autonomous  determination  upon  an  ever  in- 
creasing realm  of  events." 

It  should  be  noted  that  this  basic  actualizing  tendency  is  the  only 
motive  which  is  postulated  in  this  theoretical  system.  It  should  also  be 
noted  that  it  is  the  organism  as  a  whole,  and  only  the  organism  as  a 
whole,  which  exhibits  this  tendency.  There  are  no  homunculi,  no  other 
sources  of  energy  or  action  in  the  system.  The  self,  for  example,  is  an 
important  construct  in  our  theory,  but  the  self  does  not  "do"  anything. 
It  is  only  one  expression  of  the  general  tendency  of  the  organism  to 
behave  in  those  ways  which  maintain  and  enhance  itself. 

It  might  also  be  mentioned  that  such  concepts  of  motivation  as 
are  termed  need-reduction,  tension-reduction,  drive-reduction,  are  in- 
cluded in  this  concept.  It  also  includes,  however,  the  growth  motivations 
which  appear  to  go  beyond  these  terms:  the  seeking  of  pleasurable  ten- 
sions, the  tendency  to  be  creative,  the  tendency  to  learn  painfully  to 
walk  when  crawling  would  meet  the  same  needs  more  comfortably. 

2.  Tendency  toward  self -actualization.  Following  the  development 
of  the  self-structure,  this  general  tendency  toward   actualization   ex- 
presses itself  also  in  the  actualization  of  that  portion  of  the  experience 
of  the  organism  which  is  symbolized  in  the  self.  If  the  self  and  the  total 

Therapy,  Personality,  and  Interpersonal  Relationships  197 

experience  of  the  organism  are  relatively  congruent,  then  the  actualizing 
tendency  remains  relatively  unified.  If  self  and  experience  are  Incon- 
gnient,  then  the  general  tendency  to  actualize  the  organism  may  work 
at  cross  purposes  with  the  subsystem  of  that  motive,  the  tendency  to 
actualize  the  self. 

This  definition  will  be  better  understood  when  various  of  Its  terms — 
self,  Incongruence,  etc. — are  defined.  It  is  given  here  because  it  Is  a 
sub-aspect  of  motivation.  It  should  perhaps  be  reread  after  the  other 
terms  are  more  accurately  understood. 

3.  Experience  (noun).  This  term  is  used  to  include  all  that  Is  going 
on  within  the  envelope  of  the  organism  at  any  given  moment  wrhlch  is 
potentially  available  to  awareness.  It  includes  events  of  which  the  in- 
dividual is  unaware,  as  well  as  all  the  phenomena  which  are  in  con- 
sciousness. Thus  it  includes  the  psychological  aspects  of  hunger,  even 
though  the  individual  may  be  so  fascinated  by  his  work  or  play  that  he 
is  completely  unaware  of  the  hunger;  it  includes  the  impact  of  sights 
and  sounds  and  smells  on  the  organism,  even  though  these  are  not  in 
the  focus  of  attention.  It  includes  the  influence  of  memory  and  past 
experience,  as  these  are  active  in  the  moment,  in  restricting  or  broaden- 
ing the  meaning  given  to  various  stimuli.  It  also  includes  all  that  is 
present  in  immediate  awareness  or  consciousness.  It  does  not  Include 
such  events  as  neuron  discharges  or  changes  in  blood  sugar,  because 
these  are  not  directly  available  to  awareness.  It  is  thus  a  psychological, 
not  a  physiological  definition. 

Synonyms  are  "experiential  field,"  or  the  term  "phenomenal  field"  as 
used  by  Snygg  and  Combs,  which  also  covers  more  than  the  phenomena 
of  consciousness.  I  have  in  the  past  used  such  phrases  as  "sensory  and 
visceral  experiences"  and  "organic  experiences"  in  the  attempt  to  convey 
something  of  the  total  quality  of  this  concept. 

It  is  to  be  noted  that  experience  refers  to  the  given  moment,  not  to 
some  accumulation  of  past  experience.  It  is  believed  that  this  makes  the 
operational  definition  of  experience,  or  of  an  experience,  which  is  a 
given  segment  of  the  field,  more  possible. 

4.  Experience    (verb).   To   experience  means  simply  to  receive  in 
the  organism  the  impact  of  the  sensory  or  physiological  events  which  are 
happening  at  the  moment. 

Often  this  process  term  is  used  in  the  phrase  "to  experience  in  aware- 
ness" which  means  to  symbolize  in  some  accurate  form  at  the  conscious 
level  the  above  sensory  or  visceral  events.  Since  there  are  varying  de- 
grees of  completeness  in  symbolization,  the  phrase  is  often  "to  experience 
more  fully  in  awareness,"  thus  indicating  that  it  is  the  extension  of  this 

198  CARL   R.    ROGERS 

process  toward   more  complete   and  accurate  symbolization  to  which 
reference  is  being  made. 

5.  Peeling,  Experiencing  a  feeling.  This  is  a  term  which  has  been 
heavily  used  in  writings  on  client-centered  therapy  and  theory.  It  denotes 
an  emotionally  tinged  experience,  together  with  its  personal  meaning. 
Thus  it  includes  the  emotion  but  also  the  cognitive  content  of  the  mean- 
ing of  that  emotion  in  its  experiential  context.  It  thus  refers  to  the 
unity  of  emotion  and  cognition  as  they  are  experienced  inseparably 
in  the  moment.  It  is  perhaps  best  thought  of  as  a  brief  theme  of  ex- 
perience, carrying  with  it  the  emotional  coloring  and  the  perceived 
meaning  to  the  individual.  Examples  would  include  "I  feel  angry  at 
myself,"  "I  feel  ashamed  of  my  desires  when  I  am  with  her,"  "For  the 
first  time,  right  now,  I  feel  that  you  like  me."  This  last  is  an  example  of 
another  phenomenon  which  is  relevant  to  our  theory,  and  which  has 
been  called  experiencing  a  feeling  fully,  in  the  immediate  present.  The 
individual  is  then  congruent  in  his  experience  (of  the  feeling),  his  aware- 
ness (of  it) 3  and  his  expression  (of  it) . 

6.  Awareness,  Symbolization,  Consciousness.  These  three  terms  are 
defined  as  synonymous.  To  use  AngyaPs  expression,  consciousness  (or 
awareness)    is  the  symbolization  of  some  of  our  experience.   Aware- 
ness [is  thus  seen  as  the  symbolic  representation  (not  necessarily  in  verbal 
symbols)   of  some  portion  of  our  experience.  This  representation  may 
have  varying  degrees  of  sharpness  or  vividness,  from  a  dim  awareness 
of  something  existing  as  ground,  to  a  sharp  awareness  of  something 
which  is  in  focus  as  figure. 

7.  Availability  to  awareness.  When  an  experience  can  be  symbolized 
freely,  without  defensive  denial  and  distortion,  then  it  is  available  to 

8.  Accurate  symbolization.  The  symbols  which  constitute  our  aware- 
ness do  not  necessarily  match,  or  correspond  to,  the  "real"  experience, 
or  to  "reality."  Thus  the  psychotic  is  aware  of  (symbolizes)  electrical 
impulses  in  his  body  which  do  not  seem  in  actuality  to  exist.  I  glance 
up  quickly  and  perceive  a  plane  in  the  distance,  but  it  turns  out  to  be 
a  gnat  close  to  my  eye.  It  seems  important  to  distinguish  between  those 
awarenesses  which,  in  common-sense  terms,  are  real  or  accurate  and 
those  which  are  not.  But  how  can  this  be  conceptualized  if  we  are  trying 
to  think  rigorously? 

The  most  adequate  way  of  handling  this  predicament  seems  to  me 
to  be  to  take  the  position  of  those  who  recognize  that  all  perception 
(and  I  would  add,  all  awareness)  is  transactional  in  nature,  that  it  is 
a  construction  from  our  past  experience  and  a  hypothesis  or  prognosis 
for  the  future.  Thus  the  examples  given  are  both  hypotheses  which 

Therapy,  Personality,,  and  Interpersonal  Relationships  199 

can  be  checked.  If  I  brush  at  the  gnat  and  It  disappears,  it  increases  the 
probability  that  what  I  was  aware  of  was  a  gnat  and  not  a  plane.  If 
the  psychotic  were  able  to  permit  himself  to  check  the  electric  currents 
in  Ms  body,  and  to  see  whether  they  have  the  same  characteristics  as 
other  electric  currents,  he  would  be  checking  the  hypothesis  implicit  in 
his  awareness.  Hence  when  we  speak  of  accurate  symbolization  in  aware- 
ness, we  mean  that  the  hypotheses  implicit  in  the  awareness  will  be  borne 
out  if  tested  by  acting  on  them. 

We  are,  however,  well  over  the  border  line  of  simple  awareness  and 
into  the  realm  which  is  usually  classified  as  perception,  so  let  us  proceed 
to  a  consideration  of  that  concept. 

9.  Perceive,  Perception.   So   much  has  the  meaning  of  this  term 
changed  that  one  definition  has  been  given  as  follows:   "Perception  is 
that  which  comes  into  consciousness  wrhen  stimuli,  principally  light  or 
sound,  impinge  on  the  organism  from  the  outside"   [40,  p.  250].  Al- 
though this  seems  a  bit  too  general,  it  does  take  account  of  the  work  of 
Hebb,  Riesen,  and  others,  which  indicates  that  the  impingement  of  the 
stimuli  and  the  meaning  given  to  the  stimuli  are  inseparable  parts  of  a 
single  experience. 

For  our  own  definition  we  might  say  that  a  perception  is  a  hypothesis 
or  prognosis  for  action  which  comes  into  being  in  awareness  when 
stimuli  impinge  on  the  organism.  When  wre  perceive  "this  is  a  triangle," 
"that  is  a  tree,"  "this  person  is  my  mother,'3  it  means  that  we  are  making 
a  prediction  that  the  objects  from  which  the  stimuli  are  received  would, 
if  checked  in  other  ways,  exhibit  properties  we  have  come  to  regard, 
from  our  past  experience,  as  being  characteristic  of  triangles,  trees, 

Thus  we  might  say  that  perception  and  awareness  are  synonymous, 
perception  being  the  narrower  term,  usually  used  when  we  wish  to 
emphasize  the  importance  of  the  stimulus  in  the  process,  and  awareness 
the  broader  term,  covering  symbolizations  and  meanings  which  arise 
from  such  purely  internal  stimuli  as  memory  traces,  visceral  changes, 
and  the  like,  as  well  as  from  external  stimuli. 

To  define  perception  in  this  purely  psychological  fashion  is  not 
meant  to  deny  that  it  can  be  defined  in  physiological  fashion  by  referring 
to  the  impact  of  a  pattern  of  light  rays  upon  certain  nerve  cells,  for 
example.  For  our  purpose,  however,  the  psychological  definition  seems 
more  fruitful,  and  it  is  in  this  sense  that  the  term  will  be  used  in  our 

10.  Subceive,  Subception.  McCleary  and  Lazarus  [46]  formulated 
this  construct  to  signify  discrimination  without  awareness.  They  state 
that  "even  when  a  subject  is  unable  to  report  a  visual  discrimination  he 
is  still  able  to  make  a  stimulus  discrimination  at  some  level  below  that 

200  CARL    R.    ROGERS 

required  for  conscious  recognition.33  Thus  it  appears  that  the  organism 
can  discriminate  a  stimulus  and  its  meaning  for  the  organism  without 
utilizing  the  higher  nerve  centers  involved  in  awareness.  It  is  this  capacity 
which,  in  our  theory,  permits  the  individual  to  discriminate  an  ex- 
perience as  threatening,  without  symbolization  in  awareness  of  this 

11.  Self -experience.  This  is  a  term  coined  by  Standal   [80],   and 
defined  as  being  any  event  or  entity  in  the  phenomenal  field  discriminated 
by  the  individual  which  is  also  discriminated  as  "self,"  "me,"  "I,"  or 
related  thereto.  In  general  self-experiences  are  the  raw  material  of  which 
the  organized  self-concept  is  formed. 

12.  Self,  Concept  of  self,  Self -structure.  These  terms  refer  to  the 
organized,  consistent  conceptual  gestalt  composed  of  perceptions  of  the 
characteristics  of  the  "I"  or  "me"  and  the  perceptions  of  the  relation- 
ships of  the  "I"  or  "me33  to  others  and  to  various  aspects  of  life,  to- 
gether with  the  values  attached  to  these  perceptions.   It  is  a  gestalt 
which  is  available  to  awareness  though  not  necessarily  in  awareness. 
It  is  a  fluid  and  changing  gestalt,  a  process,  but  at  any  given  moment 
it  is  a  specific  entity  which  is  at  least  partially  definable  in  operational 
terms  by  means  of  a  Q  sort  or  other  instrument  or  measure.  The  term 
self  or  self-concept  is  more  likely  to  be  used  when  we  are  talking  of  the 
person's  view  of  himself,  self -structure  when  we  are  looking  at  this 
gestalt  from  an  external  frame  of  reference. 

13.  Ideal  self.  Ideal  self  (or  self-ideal)  is  the  term  used  to  denote 
the  self-concept  which  the  individual  would  most  like  to  possess,  upon 
which  he  places  the  highest  value  for  himself.  In  all  other  respects  it  is 
defined  in  the  same  way  as  the  self-concept. 

A  digression  on  the  case  history  of  a  construct.  Since  the  abstrac- 
tion which  we  term  the  self  is  one  of  the  central  constructs  in  our 
theory,  it  may  be  helpful  to  interpose  a  somewhat  lengthy  digression  at 
this  point  in  our  list  of  definitions  in  order  to  relate  something  of  the 
development  of  this  construct.  In  so  doing  we  will  also  be  illustrating 
the  manner  in  which  most  of  these  defined  constructs  have  come  into 
being  in  our  theory. 

Speaking  personally,  I  began  my  work  with  the  settled  notion  that 
the  "self"  was  a  vague,  ambiguous,  scientifically  meaningless  term 
which  had  gone  out  of  the  psychologist's  vocabulary  with  the  departure 
of  the  introspectionists.  Consequently  I  was  slow  in  recognizing  that 
when  clients  were  given  the  opportunity  to  express  their  problems  and 
their  attitudes  in  their  own  terms,  without  any  guidance  or  interpreta- 

Therapy,  Personality^  and  Interpersonal  Relationships  201 

tion,  they  tended  to  talk  in  terms  of  the  self.  Characteristic  expressions 
were  attitudes  such  as  these:  "I  feel  I'm  not  being  my  real  self."  "I 
wonder  who  I  am?  really."  i;I  wouldn't  want  anyone  to  know  the  real 
me.35  "I  never  had  a  chance  to  be  myself."  "It  feels  good  to  let  myself 
go  and  just  be  myself  here."  "I  think  if  I  chip  off  all  the  plaster  facade 
I've  got  a  pretty  solid  self — a  good  substantial  brick  building,  under- 
neath." It  seemed  clear  from  such  expressions  that  the  self  was  an  im- 
portant element  in  the  experience  of  the  client,  and  that  in  some  odd 
sense  his  goal  was  to  become  his  "real  self." 

Raimy  [54]  produced  a  careful  and  searching  definition  of  the  self- 
concept  which  was  helpful  in  our  thinking.  There  seemed  to  be  no 
operational  way  of  defining  it  at  that  point.  Attitudes  toward  the  self 
could  be  measured,  however,  and  Raimy  and  a  number  of  others  began 
such  research.  Self-attitudes  were  determined,  operationally,  by  the 
categorizing  of  all  self-referent  terms  in  interviews  preserved  in  verbatim 
form  by  electrical  recording.  The  categories  used  had  a  satisfactory 
degree  of  interjudge  reliability,  thus  making  them  suitable  scientific  con- 
structs for  our  work.  We  were  encouraged  to  find  that  these  self- 
referent  attitudes  altered  significantly  in  therapy  as  we  had  hypothesized 
they  would. 

As  we  focused  more  upon  the  concept  of  the  self,  clinical  experience 
again  gave  us  further  clues  as  to  its  nature.  For  example,  in  the  process 
of  change  which  appeared  to  occur  in  therapy,  it  was  not  at  all  un- 
common to  find  violent  fluctuation  in  the  concept  of  the  self.  A  client, 
during  a  given  interview,  would  come  to  experience  himself  quite 
positively.  He  felt  he  was  worthwhile,  that  he  could  meet  life  with  the 
capacities  he  possessed,  and  that  he  was  experiencing  a  quiet  confidence. 
Three  days  later  he  might  return  with  a  completely  reversed  conception 
of  himself.  The  same  evidence  now  proved  an  opposite  point.  The  posi- 
tive new  choice  he  had  made  now  was  an  instance  of  silly  immaturity; 
the  valid  feelings  courageously  expressed  to  his  colleagues  now  were 
clearly  inadequate.  Often  such  a  client  could  date,  to  the  moment,  the 
point  at  which,  following  some  very  minor  incident,  the  balance  was 
upset,  and  his  picture  of  himself  had  undergone  a  complete  flip-flop. 
During  the  interview  it  might  as  suddenly  reverse  itself  again. 

Consideration  of  this  phenomenon  made  it  clear  that  we  were  not 
dealing  with  an  entity  of  slow  accretion,  of  step-by-step  learning,  of 
thousands  of  unidirectional  conditionings.  These  might  all  be  involved, 
but  the  product  was  clearly  a  gestalt,  a  configuration  in  which  the  alter- 
ation of  one  minor  aspect  could  completely  alter  the  whole  pattern.  One 
was  forcibly  reminded  of  the  favorite  textbook  illustration  of  a  gestalt, 
the  double  picture  of  the  old  hag  and  the  young  woman.  Looked  at  with 
one  mind  set,  the  picture  is  clearly  that  of  an  ugly  old  woman.  The 



slightest  change,  and  the  whole  becomes  a  portrait  of  an  attractive  girl. 
So  with  our  clients.  The  self-concept  was  clearly  configurational  in 

Our  clinical  experience  gave  us  another  clue  to  the  manner  in  which 
the  self  functioned.  The  conventional  concept  of  repression  as  having  to 
do  with  forbidden  or  socially  taboo  impulses  had  been  recognized  as  in- 
adequate to  fit  the  facts.  Often  the  most  deeply  denied  impulses  and 
feelings  were  positive  feelings  of  love,  or  tenderness,  or  confidence  in 
self.  How  could  one  explain  the  puzzling  conglomeration  of  experience 
which  seemingly  could  not  be  permitted  in  awareness?  Gradually  it  was 
recognized  that  the  important  principle  was  one  of  consistency  with  the 
self.  Experiences  which  were  incongruent  with  the  individual's  concept 
of  himself  tended  to  be  denied  to  awareness,  whatever  their  social  char- 
acter. We  began  to  see  the  self  as  a  criterion  by  which  the  organism 
screened  out  experiences  which  could  not  comfortably  be  permitted  in 
consciousness.  Lecky's  little  posthumous  book  [43]  reinforced  this  line 
of  thought.  We  also  began  to  understand  other  functions  of  the  self  in  its 
regulatory  influence  on  behavior,  and  the  like. 

At  about  this  juncture  Stephenson's  Q  technique  [81]  opened  up 
the  possibility  of  an  operational  definition  of  the  self-concept.  Im- 
mediately, research  burgeoned.  Though  we  feel  it  has  barely  made  a 
start  in  exploiting  the  possible  testing  of  hypotheses,  there  have  already 
been  measurements  and  predictions  regarding  the  self  as  of  this  moment, 
the  self  in  the  past,  "myself  as  I  am  with  my  mother,"  "the  self  I  would 
like  to  be,"  etc.  Probably  the  most  sophisticated  and  significant  of  these 
studies  is  that  completed  by  ChodorkofT  [10],  in  which  his  hypothesis, 
stated  informally,  is  as  follows:  that  the  greater  the  agreement  between 
the  individual's  self-description  and  an  objective  description  of  him, 
the  less  perceptual  defensiveness  he  will  show,  and  the  more  adequate 
will  be  his  personal  adjustment.  This  hypothesis  is  upheld  and  tends  to 
confirm  some  important  aspects  of  our  theory.  In  general  the  various 
investigations  have  agreed  in  indicating  that  the  self-concept  is  an  im- 
portant variable  in  personality  dynamics  and  that  change  in  the  self  is 
one  of  the  most  marked  and  significant  changes  occurring  in  therapy. 

It  should  be  recognized  that  any  construct  is  a  more  or  less  arbitrary 
abstraction  from  experience.  Thus  the  self  could  be  defined  in  many  dif- 
ferent ways.  Hilgard,  for  example  [34],  has  proposed  that  it  be  defined 
in  such  a  way  as  to  include  unconscious  material,  not  available  to  aware- 
ness, as  well  as  conscious  material.  Although  we  recognize  that  this  is 
certainly  a  legitimate  way  of  abstracting  from  the  phenomena,  we  be- 
lieve it  is  not  a  useful  way  because  it  produces  a  concept  which  cannot 
at  this  point  be  given  operational  definition.  One  cannot  obtain  sufficient 
agreement  as  to  the  content  of  the  individual's  unconscious  to  make 

Therapy,  Personality,  and  Interpersonal  Relationships  203 

research  possible.  Hence  we  believe  that  it  is  more  fruitful  to  define  the 
self-concept  as  a  gestalt  which  is  available  to  awareness.  This  has  per- 
mitted and  encouraged  a  flood  of  important  research. 

At  all  times,  however,  we  endeavor  to  keep  in  the  forefront  of  our 
thinking  the  fact  that  each  definition  is  no  more  than  an  abstraction 
and  that  the  same  phenomena  might  be  abstracted  in  a  different  fashion. 
One  of  our  group  is  working  on  a  definition  of  self  which  would  give 
more  emphasis  to  its  process  nature.  Others  have  felt  that  a  plural 
definition,  indicating  many  specific  selves  in  each  of  various  life  contexts, 
would  be  more  fruitful,  and  this  way  of  thinking  has  been  embodied  in, 
for  example,  XunnaUy's  [50]  research.  So  the  search  continues  for  a  more 
adequate  conceptualization  of  this  area  of  our  therapeutic  experience 
and  for  more  adequate  technical  means  of  providing  operational  defini- 
tions for  the  concepts  which  are  formulated. 

This  concludes  our  interruption  of  the  list  of  definitions.  It  is  hoped 
that  this  one  example  will  give  an  indication  of  the  way  in  which  many 
of  our  basic  constructs  have  developed — not  only  the  self-concept  but 
the  constructs  of  congruence,  incongraence,  defensiveness,  unconditional 
positive  regard,  locus  of  evaluation,  and  the  like.  Although  the  process 
has  been  irregular,  it  has  tended  to  include  clinical  observation,  initial 
conceptualization,  initial  crude  research  to  test  some  of  the  hypotheses 
involved,  further  clinical  observation,  more  rigorous  formulation  of  the 
construct  and  its  functional  relationships,  more  refined  operational 
definitions  of  the  construct,  more  conclusive  research. 

14.  Incongruence  between  self  and  experience.  In  a  manner  which 
will  be  described  in  the  theory  of  personality  a  discrepancy  frequently 
develops  between  the  self  as  perceived,  and  the  actual  experience  of  the 
organism.   Thus  the  individual  may  perceive  himself  as  having  char- 
acteristics a,  by  and  c3  and  experiencing  feelings  x,  y\  and  z.  An  accurate 
symbolization  of  his  experience  would,  however,  indicate  characteristics 
c,  d}  and  e,  and  feelings  v,  w3  x.  When  such  a  discrepancy  exists,  the 
state  is  one  of  incongruence  between  self  and  experience.  This  state  is 
one  of  tension  and  internal  confusion,  since  in  some  respects  the  in- 
dividual's behavior  will  be  regulated  by  the  actualizing  tendency,  and  in 
other  respects  by  the  self-actualizing  tendency,  thus  producing  discordant 
or  incomprehensible  behaviors.  What  is  commonly  called  neurotic  be- 
havior is  one  example,  the  neurotic  behavior  being  the  product  of  the 
actualizing  tendency,  whereas  in  other  respects  the  individual  is  actualiz- 
ing the  self.  Thus  the  neurotic  behavior  is  incomprehensible  to  the  in- 
dividual himself,  since  it  is  at  variance  with  what  he  consciously  "wants" 
to  do,  which  is  to  actualize  a  self  no  longer  congruent  with  experience. 

15.  Vulnerability.  Vulnerability  is  the  term  used  to  refer  to  the 

204  CARL   R.    ROGERS 

state  of  incongruence  between  self  and  experience,  when  it  is  desired 
to  emphasize  the  potentialities  of  this  state  for  creating  psychological  dis- 
organization. When  incongruence  exists,  and  the  individual  is  unaware 
of  it,  then  he  is  potentially  vulnerable  to  anxiety,  threat,  and  disorganiza- 
tion. If  a  significant  new  experience  demonstrates  the  discrepancy  so 
clearly  that  it  must  be  consciously  perceived,  then  the  individual  will  be 
threatened,  and  his  concept  of  self  disorganized  by  this  contradictory  and 
unassimilable  experience. 

16.  Anxiety.  Anxiety  is  phenomenologically  a  state  of  uneasiness 
or  tension  whose  cause  is  unknown.  From  an  external  frame  of  reference, 
anxiety  is  a  state  in  which  the  incongruence  between  the  concept  of  self 
and  the  total  experience  of  the  individual  is  approaching  symbolization 
in  awareness.  When  experience  is  obviously  discrepant  from  the  self- 
concept,  a  defensive  response  to  threat  becomes  increasingly  difficult. 
Anxiety  is  the  response  of  the  organism  to  the  "subception"  that  such 
discrepancy  may  enter  awareness,  thus  forcing  a  change  in  the  self- 

17.  Threat.  Threat  is  the  state  which  exists  when  an  experience  is 
perceived  or  anticipated  (subceived)  as  incongruent  with  the  structure 
of  the  self.   It  may  be  regarded   as  an   external  view   of  the  same 
phenomenon  which,  from  the  internal  frame  of  reference,  is  anxiety. 

18.  Psychological  maladjustment.  Psychological  maladjustment  exists 
when  the  organism  denies  to  awareness,  or  distorts  in  awareness,  sig- 
nificant experiences,  which  consequently  are  not  accurately  symbolized 
and  organized  into  the  gestalt  of  the  self-structure,  thus  creating  an  in- 
congruence between  self  and  experience. 

It  may  help  to  clarify  this  basic  concept  of  incongruence  if  we  recog- 
nize that  several  of  the  terms  we  are  defining  are  simply  different 
vantage  points  for  viewing  this  phenomenon.  If  an  individual  is  in  a  state 
of  incongruence  between  self  and  experience  and  we  are  looking  at 
him  from  an  external  point  of  view  we  see  him  as  vulnerable  (if  he  is 
unaware  of  the  discrepancy),  or  threatened  (if  he  has  some  awareness 
of  it) .  If  we  are  viewing  him  from  a  social  point  of  view,  then  this  in- 
congruence is  psychological  maladjustment.  If  the  individual  is  viewing 
himself,  he  may  even  see  himself  as  adjusted  (if  he  has  no  awareness  of 
the  discrepancy)  or  anxious  (if  he  dimly  subceives  it)  or  threatened 
or  disorganized  (if  the  discrepancy  has  forced  itself  upon  his  awareness). 

19.  Defense,  Defensiveness.  Defense  is  the  behavioral  response  of  the 
organism  to  threat,  the  goal  of  which  is  the  maintenance  of  the  current 
structure  of  the  self.  This  goal  is  achieved  by  the  perceptual  distortion 
of  the  experience  in  awareness,  in  such  a  way  as  to  reduce  the  incongruity 
between  the  experience  and  the  structure  of  the  self,  or  by  the  denial 

Therapy,  Personality,  and  Interpersonal  Relationships  205 

to  awareness  of  an  experience,  thus  denying  any  threat  to  the  self.  De- 
fensiveness  is  the  term  denoting  a  state  in  which  the  behaviors  are  of 
the  sort  described. 

20.  Distortion  in  awareness^  Denial  to  awareness.  It  is  an  observed 
phenomenon  that  material  which  is  significantly  inconsistent  with  the 
concept  of  self  cannot  be  directly  and  freely  admitted  to  awareness.  To 
explain  this  the  construct  of  denial  or  distortion  has  been  developed. 
When  an  experience  is  dimly  perceived  (or  "subceived"  is  perhaps  the 
better  term)  as  being  incongraent  with  the  self-structure,  the  organism 
appears  to  react  with  a  distortion  of  the  meaning  of  the  experience, 
(making  it  consistent  with  the  self)   or  with  a  denial  of  the  existence 
of  the  experience,  in  order  to  preserve  the  self-structure  from  threat.  It  is 
perhaps  most  vividly  illustrated  in  those  occasional  moments  in  therapy 
when  the  therapist's  response,  correctly  heard  and  understood,  would 
mean  that  the  client  would  necessarily  perceive  openly  a  serious  in- 
consistency between  his  self-concept  and  a  given  experience.  In  such  a 
case,  the  client  may  respond,  "I  can  hear  the  words  you  say,  and  I  know 
I  should  understand  them,  but  I  just  can't  make  them  convey  any  mean- 
ing to  me."  Here  the  relationship  is  too  good  for  the  meaning  to  be 
distorted  by  rationalization,  the  meaning  too  threatening  to  be  received. 
Hence  the  organism  denies  that  there  is  meaning  in  the  communication. 
Such  outright   denial   of  experience   is  much  less  common  than   the 
phenomenon   of  distortion.   Thus  if  the  concept  of  self  includes  the 
characteristic  "I  am  a  poor  student"  the  experience  of  receiving  a  high 
grade  can  be  easily  be  distorted  to  make  it  congruent  with  the  self  by 
perceiving  in  it  such  meanings  as,  "That  professor  is  a  fool";  "It  was 
just  luck53 ;  etc. 

21.  Intensionality.  This  term  is  taken  from  general  semantics.   If 
the  person  is  reacting  or  perceiving  in  an  intensional  fashion  he  tends 
to  see  experience  in  absolute  and  unconditional  terms,  to  overgeneralize, 
to  be  dominated  by  concept  or  belief,  to  fail  to  anchor  his  reactions  in 
space  and  time,  to  confuse  fact  and  evaluation,  to  rely  upon  abstractions 
rather  than  upon  reality-testing.  This  term  covers  the  frequently  used 
concept  of  rigidity  but  includes  perhaps  a  wider  variety  of  behaviors 
than  are  generally  thought  of  as  constituting  rigidity. 

It  will  perhaps  be  evident  that  this  cluster  of  definitions  all  have  to 
do  with  the  organism's  response  to  threat.  Defense  is  the  most  general 
term:  distortion  and  denial  are  the  mechanisms  of  defense;  intensionality 
is  a  term  which  covers  the  characteristics  of  the  behavior  of  the  in- 
dividual who  is  in  a  defensive  state. 

22.  Congruence,  Congruence  of  self  and  experience.  This  is  a  basic 
concept  which  has  grown  out  of  therapeutic  experience,  in  which  the 

206  CARL    R.    ROGERS 

individual  appears  to  be  revising  his  concept  of  self  to  bring  it  into  con- 
gruence with  his  experience,  accurately  symbolized.  Thus  he  discovers 
that  one  aspect  of  his  experience  if  accurately  symbolized,  would  be 
hatred  for  his  father;  another  would  be  strong  homosexual  desires.  He 
reorganizes  the  concept  he  holds  of  himself  to  include  these  char- 
acteristics, which  would  previously  have  been  inconsistent  with  self. 

Thus  when  self-experiences  are  accurately  symbolized,  and  are  in- 
cluded in  the  self-concept  in  this  accurately  symbolized  form,  then  the 
state  is  one  of  congruence  of  self  and  experience.  If  this  were  completely 
true  of  all  self-experiences,  the  individual  would  be  a  fully  functioning 
person,  as  will  be  made  more  clear  in  the  section  devoted  to  this  aspect 
of  our  theory.  If  it  is  true  of  some  specific  aspect  of  experience,  such  as 
the  individual's  experience  in  a  given  relationship  or  in  a  given  moment 
of  time,  then  we  can  say  that  the  individual  is  to  this  degree  in  a  state 
of  congruence.  Other  terms  which  are  in  a  general  way  synonymous 
are  these :  integrated,  whole,  genuine. 

23.  Openness  to   experience.   When  the  individual  is  in  no  way 
threatened,  then  he  is  open  to  his  experience.  To  be  open  to  experience 
is  the  polar  opposite  of  defensiveness.  The  term  may  be  used  in  regard  to 
some  area  of  experience  or  in  regard  to  the  total  experience  of  the 
organism.   It  signifies  that  every  stimulus,  whether  originating  within 
the  organism  or  in  the  environment,  is  freely  relayed  through  the  nervous 
system  without  being  distorted  or  channeled  off  by  any  defensive  mecha- 
nism. There  is  no  need  of  the  mechanism  of  "subception"  whereby  the 
organism  is  forewarned  of  experiences  threatening  to  the  self.  On  the 
contrary,  whether  the  stimulus  is  the  impact  of  a  configuration  of  form, 
color,  or  sound  in  the  environment  on  the  sensory  nerves,  or  a  memory 
trace  from  the  past,  or  a  visceral  sensation  of  fear,  pleasure,  or  disgust, 
it  is  completely  available  to  the  individual's  awareness.  In  the  hypo- 
thetical person  who  is  completely  open  to  his  experience,  his  concept  of 
self  would  be  a  symbolization  in  awareness  which  would  be  completely 
congruent  with  his  experience.  There  would,  therefore,  be  no  possibility 
of  threat. 

24.  Psychological    adjustment.    Optimal    psychological    adjustment 
exists  when  the  concept  of  the  self  is  such  that  all  experiences  are  or 
may  be  assimiliated  on  a  symbolic  level  into  the  gestalt  of  the  self- 
structure.  Optimal  psychological  adjustment  is  thus  synonymous  with 
complete  congruence  of  self  and  experience,  or  complete  openness  to 
experience.  On  the  practical  level,  improvement  in  psychological  ad- 
justment is  equivalent  to  progress  toward  this  end  point. 

25.  Extensionality.  This  term  is  taken  from  general  semantics.   If 
the  person  is  reacting  or  perceiving  in  an  extensional  manner  he  tends 
to  see  experience  in  limited,  differentiated  terms,  to  be  aware  of  the 

Therapy,  Personality,  and  Interpersonal  Relationships  207 

space-time  anchorage  of  facts,  to  be  dominated  by  facts,  not  by  concepts, 
to  evaluate  in  multiple  ways,  to  be  aware  of  different  levels  of  abstraction, 
to  test  his  inferences  and  abstractions  against  reality. 

26.  Mature,  Maturity.  The  individual  exhibits  mature  behavior  when 
he  perceives  realistically  and  in  an  extensional  manner,  is  not  defensive, 
accepts  the  responsibility  of  being  different  from  others,  accepts  re- 
sponsibility for  his  own  behavior,  evaluates  experience  in  terms  of  the 
evidence  coming  from  his  own  senses,  changes  his  evaluation  of  ex- 
perience only  on  the  basis  of  new  evidence,  accepts  others  as  unique  in- 
dividuals different  from  himself,  prizes  himself,  and  prizes  others.  (If  his 
behavior  has  these  characteristics,  then  there  will  automatically  follow 
all  the  types  of  behavior  which  are  more  popularly  thought  of  as  con- 
stituting psychological  maturity. ) 

These  last  five  definitions  form  a  cluster  which  grows  out  of  the  con- 
cept of  congruence.  Congruence  is  the  term  which  defines  the  state. 
Openness  to  experience  is  the  way  an  internally  congruent  individual 
meets  new  experience.  Psychological  adjustment  is  congruence  as  viewed 
from  a  social  point  of  view.  Extensional  is  the  term  which  describes  the 
specific  types  of  behavior  of  a  congruent  individual.  Maturity  is  a 
broader  term  describing  the  personality  characteristics  and  behavior  of 
a  person  who  is,  in  general,  congruent. 

The  concepts  in  the  group  of  definitions  which  follow  have  all  been 
developed  and  formulated  by  Standal  [80],  and  have  taken  the  place 
of  a  number  of  less  satisfactory  and  less  rigorously  defined  constructs. 
Essentially  this  group  has  to  do  with  the  concept  of  positive  regard,  but 
since  all  transactions  relative  to  this  construct  take  place  in  relationships, 
a  definition  of  psychological  contact,  or  minimal  relationship,  is  set  down 

27.  Contact.  Two  persons  are  in  psychological  contact,  or  have  the 
minimum  essential  of  a  relationship,  when  each  makes  a  perceived  or 
subceived  difference  in  the  experiential  field  of  the  other. 

This  construct  was  first  given  the  label  of  "relationship"  but  it  was 
found  that  this  led  to  much  misunderstanding,  for  it  was  often  under- 
stood to  represent  the  depth  and  quality  of  a  good  relationship,  or  a 
therapeutic  relationship.  The  present  term  has  been  chosen  to  signify 
more  clearly  that  this  is  the  least  or  minimum  experience  which  could 
be  called  a  relationship.  If  more  than  this  simple  contact  between  two 
persons  is  intended,  then  the  additional  characteristics  of  that  contact 
are  specified  in  the  theory. 

28.  Positive  regard.  If  the  perception  by  me  of  some  self-experience 
in  another  makes  a  positive  difference  in  my  experiential  field,  then  I 

208  CARL   R.   ROGERS 

am  experiencing  positive  regard  for  that  individual.  In  general,  positive 
regard  is  defined  as  including  such  attitudes  as  warmth,  liking,  respect, 
sympathy,  acceptance.  To  perceive  oneself  as  receiving  positive  regard  is 
to  experience  oneself  as  making  a  positive  difference  in  the  experiential 
field  of  another. 

29.  Need  for  positive  regard.  It  is  postulated  by  Standal  that  a  basic 
need  for  positive  regard,  as  defined  above,  is  a  secondary  or  learned 
need,  commonly  developed  in  early  infancy.  Some  writers  have  looked 
upon  the  infant's  need  for  love  and  affection  as  an  inherent  or  instinctive 
need.  Standal  is  probably  on  safer  ground  in  regarding  it  as  a  learned 
need.  By  terming  it  the  need  for  positive  regard,  he  has,  it  is  believed, 
selected  out  the  significant  psychological  variable  from  the  broader  terms 
usually  used. 

30.  Unconditional  positive  regard.  Here  is  one  of  the  key  constructs 
of  the  theory,  which  may  be  defined  in  these  terms:  if  the  self-experiences 
of  another  are  perceived  by  me  in  such  a  way  that  no  self-experience  can 
be  discriminated  as  more  or  less  worthy  of  positive  regard  than  any  other, 
then  I  am  experiencing  unconditional  positive  regard  for  this  individual. 
To  perceive  oneself  as  receiving  unconditional  positive  regard  is  to  per- 
ceive that  of  one's  self-experiences  none  can  be  discriminated  by  the 
other  individual  as  more  or  less  worthy  of  positive  regard. 

Putting  this  in  simpler  terms,  to  feel  unconditional  positive  regard 
toward  another  is  to^prize''  him  (to  use  Dewey's  term,  recently  used 
in  this  sense  by  Butler) .  This  means  to  value  the  person,  irrespective  of 
the  differential  values  which  one  might  place  on  his  specific  behaviors. 
A  parent  "prizes"  his  child,  though  he  may  not  value  equally  all  of  his 
behaviors.  Acceptance  is  another  term  which  has  been  frequently  used 
to  convey  this  meaning,  but  it  perhaps  carries  more  misleading  con- 
notations than  the  phrase  which  Standal  has  coined.  In  general,  how- 
ever, acceptance  and  prizing  are  synonymous  with  unconditional  positive 

This  construct  has  been  developed  out  of  the  experiences  of  therapy, 
where  it  appears  that  one  of  the  potent  elements  in  the  relationship  is 
that  the  therapist  "prizes"  the  whole  person  of  the  client.  It  is  the  fact 
that  he  feels  and  shows  an  unconditional  positive  regard  toward  the  ex- 
periences of  which  the  client  is  frightened  or  ashamed,  as  well  as  to- 
ward the  experiences  with  which  the  client  is  pleased  or  satisfied,  that 
seems  effective  in  bringing  about  change.  Gradually  the  client  can  feel 
more  acceptance  of  all  of  his  own  experiences,  and  this  makes  him  again 
more  of  a  whole  or  congruent  person,  able  to  function  effectively.  This 
clinical  explanation  will,  it  is  hoped,  help  to  illuminate  the  meaning 
contained  in  the  rigorous  definition. 

31.  Regard  complex.  The  regard  complex  is  a  construct  defined  by 

Therapy,  Personality,  and  Interpersonal  Relationships  209 

Standal  as  all  those  self-experiences,  together  with  their  interrelation- 
ships, which  the  individual  discriminates  as  being  related  to  the  positive 
regard  of  a  particular  social  other. 

This  construct  is  intended  to  emphasize  the  gestalt  nature  of  trans- 
actions involving  positive  or  negative  regard,  and  their  potency.  Thus, 
for  example,  if  a  parent  shows  positive  regard  to  a  child  in  relationship 
to  a  specific  behavior,  this  tends  to  strengthen  the  whole  pattern  of 
positive  regard  which  has  previously  been  experienced  as  coming  from 
that  parent.  Likewise  specific  negative  regard  from  this  parent  tends  to 
weaken  the  whole  configuration  of  positive  regard. 

32.  Positive  self -regard.  This  term  is  used  to  denote  a  positive  regard 
satisfaction  which  has  become  associated  with  a  particular  self-experience 
or  a  group  of  self-experiences,  in  which  this  satisfaction  is  independent 
of  positive  regard  transactions  with  social  others.  Though  it  appears  that 
positive  regard  must  first  be  experienced  from  others,  this  results  in  a 
positive  attitude  toward  self  which  is  no  longer  directly  dependent  on 
the   attitudes  of  others.  The  individual,  in  effect,  becomes  his  own 
significant  social  other. 

33.  Need  for  self -regard.  It  is  postulated  that  a  need  for  positive 
self-regard  is  a  secondary  or  learned  need,  related  to  the  satisfaction  of 
the  need  for  positive  regard  by  others. 

34.  Unconditional  self-regard.  When  the  individual  perceives  him- 
self in  such  a  way  that  no  self-experience  can  be  discriminated  as  more  or 
less  worthy  of  positive  regard  than  any  other,  then  he  is  experiencing 
unconditional  positive  self-regard. 

35.  Conditions  of  worth.  The  self-structure  is  characterized  by  a 
condition  of  worth  when  a  self-experience  or  set  of  related  self-ex- 
periences is  either  avoided  or  sought  solely  because  the  individual  dis- 
criminates it  as  being  less  or  more  worthy  of  self-regard. 

This  important  construct  has  been  developed  by  Standal  to  take  the 
place  of  "introjected  value,"  which  was  a  less  exact  concept  used  in 
earlier  formulations.  A  condition  of  worth  arises  when  the  positive 
regard  of  a  significant  other  is  conditional,  when  the  individual  feels 
that  in  some  respects  he  is  prized  and  in  others  not.  Gradually  this  same 
attitude  is  assimilated  into  his  own  self-regard  complex,  and  he  values 
an  experience  positively  or  negatively  solely  because  of  these  conditions 
of  worth  which  he  has  taken  over  from  others,  not  because  the  ex- 
perience enhances  or  fails  to  enhance  his  organism. 

It  is  this  last  phrase  which  deserves  special  note.  When  the  individual 
has  experienced  unconditional  positive  regard,  then  a  new  experience 
is  valued  or  not,  depending  on  its  effectiveness  in  maintaining  or  en- 

210  CARL   R.    ROGERS 

hancing  the  organism.  But  if  a  value  is  "introjected"  from  a  significant 
other,  then  this  condition  of  worth  is  applied  to  an  experience  quite 
without  reference  to  the  extent  to  which  it  maintains  or  enhances  the 
organism.  It  is  an  important  specific  instance  of  inaccurate  symboliza- 
tion,  the  individual  valuing  an  experience  positively  or  negatively,  as  if 
in  relation  to  the  criterion  of  the  actualizing  tendency,  but  not  actually 
in  relation  to  it.  An  experience  may  be  perceived  as  organismically 
satisfying,  when  in  fact  this  is  not  true.  Thus  a  condition  of  worth,  be- 
cause it  disturbs  the  valuing  process,  prevents  the  individual  from 
functioning  freely  and  with  maximum  effectiveness. 

36.  Locus  of  evaluation.  This  term  is  used  to  indicate  the  source 
of  evidence  as  to  values.  Thus  an  internal  locus  of  evaluation,  within  the 
individual  himself,  means  that  he  is  the  center  of  the  valuing  process, 
the  evidence  being  supplied  by  his  own  senses.   When  the  locus,  of 
evaluation  resides  in  others,  their  judgment  as  to  the  value  of  an  object 
or  experience  becomes  the  criterion  of  value  for  the  individual. 

37.  Organismic  valuing  process.  This  concept  describes  an  ongoing- 
process  in  which  values  are  never  fixed  or  rigid,  but  experiences  are 
being  accurately  symbolized  and  continually  and  freshly  valued  in  terms 
of  the  satisfactions  organismically  experienced;  the  organism  experiences 
satisfaction  in  those  stimuli  or  behaviors  which  maintain  and  enhance 
the  organism  and  the  self,  both  in  the  immediate  present  and  in  the  long 
range.  The  actualizing  tendency  is  thus  the  criterion.  The  simplest  ex- 
ample is  the  infant  who  at  one  moment  values  food,  and  when  satiated, 
is  disgusted  with  it;  at  one  moment  values  stimulation,  and  soon  after, 
values  only  rest;  who  finds  satisfying  that  diet  which  in  the  long  run  most 
enhances  his  development. 

38.  Internal  frame  of  reference.  This  is  all  of  the  realm  of  experience 
which  is  available  to  the  awareness  of  the  individual  at  a  given  moment. 
It  includes  the  full  range  of  sensations,  perceptions,  meanings,   and 
memories,  which  are  available  to  consciousness. 

The  internal  frame  of  reference  is  the  subjective  world  of  the  in- 
dividual. Only  he  knows  it  fully.  It  can  never  be  known  to  another  ex- 
cept through  empathic  inference  and  then  can  never  be  perfectly  known. 

39.  Empathy.  The  state  of  empathy,  or  being  empathic,  is  to  per- 
ceive the  internal  frame  of  reference  of  another  with  accuracy,  and  with 
the  emotional  components  and  meanings  which  pertain  thereto,  as  if  one 
were  the  other  person,  but  without  ever  losing  the  ccas  if  condition. 
Thus  it  means  to  sense  the  hurt  or  the  pleasure  of  another  as  he  senses 
it,  and  to  perceive  the  causes  thereof  as  he  perceives  them,  but  without 

Therapy,  Personality,  and  Interpersonal  Relationships  211 

ever  losing  the  recognition  that  it  is  as  if  I  were  hurt  or  pleased,  etc. 
If  this  "as  if"  quality  is  lost,  then  the  state  is  one  of  Identification. 

40.  External  frame  of  reference.  To  perceive  solely  from  one's  own 
subjective  internal  frame  of  reference  without  empathizing  with  the  ob- 
served person  or  object,  is  to  perceive  from  an  external  frame  of 
reference.  The  "empty  organism"  school  of  thought  in  psychology  is  an 
example  of  this.  Thus  the  observer  says  that  an  animal  has  been 
stimulated  when  the  animal  has  been  exposed  to  a  condition  which,  in 
the  observer's  subjective  frame  of  reference,  is  a  stimulus.  There  is  no 
attempt  to  understand,  empathically,  whether  this  is  a  stimulus  in  the 
animal's  experiential  field.  Likewise  the  observer  reports  that  the  animal 
emits  a  response  when  a  phenomenon  occurs  which,  in  the  observer's 
subjective  field,  is  a  response. 

We  generally  regard  all  "objects"  (stones,  trees,  or  abstractions)  from 
this  external  frame  of  reference  since  we  assume  that  they  have  no  "ex- 
perience" with  which  we  can  empathize.  The  other  side  of*  this  coin  is 
that  anything  perceived  from  an  external  frame  of  reference  (whether 
an  inanimate  thing,  an  animal,  or  a  person)  becomes  for  us  an  "object" 
because  no  empathic  inferences  are  made. 

This  cluster  of  three  ways  of  knowing  deserves  some  further  com- 
ment. In  so  far  as  we  are  considering  knowledge  of  human  beings  we 
might  say  that  these  ways  of  knowing  exist  on  a  continuum.  They  range 
from  one's  own  complete  subjectivity  in  one's  own  internal  frame  of 
reference  to  one's  own  complete  subjectivity  about  another  (the  external 
frame  of  reference).  In  between  lies  the  range  of  empathic  inference 
regarding  the  subjective  field  of  another. 

Each  of  these  ways  of  knowing  is  essentially  a  formulation  of  hypoth- 
eses. The  differences  lie  in  the  way  the  hypotheses  are  checked.  In  my 
own  internal  frame  of  reference  if  I  experience  love  or  hate,  enjoyment 
or  dislike,  interest  or  boredom,  belief  or  disbelief,  the  only  way  I  can 
check  these  hypotheses  of  experience  is  by  further  focusing  on  my 
experience.  Do  I  really  love  him?  Am  I  really  enjoying  this?  Do  I  really 
believe  this?  are  questions  which  can  only  be  answered  by  checking 
with  my  own  organism.  (If  I  try  to  find  out  whether  I  really  love  him 
by  checking  with  others,  then  I  am  observing  myself  as  an  object,  am 
viewing  myself  from  an  external  frame  of  reference. ) 

Although  in  the  last  analysis  each  individual  lives  in  and  by  his 
own  subjective  knowledge,  this  is  not  regarded  socially  as  "knowledge" 
and  certainly  not  as  scientific  knowledge. 

Knowledge  which  has  any  "certainty,"  in  the  social  sense,  involves 
the  use  of  empathic  inference  as  a  means  of  checking,  but  the  direction 
of  that  empathy  differs.  When  the  experience  of  empathic  understanding 
is  used  as  a  source  of  knowledge,  one  checks  one's  empathic  inferences 

212  CARL    R.    ROGERS 

with  the  subject,  thus  verifying  or  disproving  the  inferences  and  hypoth- 
eses implicit  in  such  empathy.  It  is  this  way  of  knowing  which  we  have 
found  so  fruitful  in  therapy.  Utilizing  empathic  inference  to  the  fullest, 
the  knowledge  thus  gained  of  the  client's  subjective  world  has  led  to 
understanding  the  basis  of  his  behavior  and  the  process  of  personality 

In  knowing  a  person  or  an  object  from  the  external  frame  of  refer- 
ence, our  implicit  hypotheses  are  checked  with  other  people,  but  not  with 
the  subject  of  our  concern.  Thus  a  rigorous  behaviorist  believes  that  S 
is  a  stimulus  for  his  experimental  animal  and  R  is  a  response,  because 
his  colleagues  and  even  the  man  in  the  street  agree  with  him  and  re- 
gard S  and  R  in  the  same  way.  His  empathic  inferences  are  made 
in  regard  to  the  internal  frame  of  reference  of  his  colleagues,  rather  than 
in  regard  to  the  internal  frame  of  reference  of  the  animal. 

Science  involves  taking  an  external  frame  of  reference,  in  which  we 
check  our  hypotheses  basically  through  empathic  inferences  as  to  the 
internal  frame  of  reference  of  our  colleagues.  They  perform  the  same 
operations  we  have  (either  actually  or  through  symbolic  representation), 
and  if  they  perceive  the  same  events  and  meanings,  then  we  regard  our 
hypotheses  as  confirmed. 

The  reason  for  thus  elaborating  the  different  ways  of  knowing  is  that 
it  seems  to  us  that  all  ways  of  knowing  have  their  usefulness,  and  that 
confusion  arises  only  when  one  is  not  clear  as  to  the  type  of  knowledge 
which  is  being  specified.  Thus  in  the  theory  of  therapy  which  follows  one 
will  find  certain  conditions  of  therapy  specified  as  subjective  experiencing 
states,  another  as  an  empathic  knowledge  of  the  client,  and  yet  the 
scientific  checking  of  the  hypotheses  of  the  theory  can  only  be  done  from 
an  external  frame  of  reference. 


This  theory  is  of  the  if-then  variety.  If  certain  conditions  exist 
(independent  variables),  then  a  process  (dependent  variable)  will 
occur  which  includes  certain  characteristic  elements.  If  this  process 
(now  the  independent  variable)  occurs,  then  certain  personality  and 
behavioral  changes  (dependent  variables)  will  occur.  This  will  be  made 

In  this  and  the  following  sections  the  formal  statement  of  the 
theory  is  given  briefly,  in  smaller  type.  The  italicized  terms  or 
phrases  in  these  formal  statements  have  been  defined  in  the  previous 
section  and  are  to  be  understood  as  defined.  The  remaining  paragraphs 
are  explanatory  and  do  not  follow  the  rigorous  pattern  of  the  formal 

Therapy,  Personality,  and  Interpersonal  Relationships  213 

A.  Conditions  of  the  Therapeutic  Process 

For  therapy  to  occur  it  is  necessary  that  these  conditions  exist. 

1.  That  two  persons  are  in  contact. 

2.  That  the  first  person,  whom  we  shall  term  the  client,  is  in  a  state  of 
incongruence,  being  vulnerable,  or  anxious. 

3.  That  the  second  person,  whom  we  shall  term  the  therapist,  is  con- 
gruent in  the  relationship. 

4.  That  the  therapist  is  experiencing  unconditional  positive  regard  to- 
ward the  client. 

5.  That  the  therapist  is  experiencing  an  empathic  understanding  of  the 
client's  internal  frame  of  reference. 

6.  That  the  client  perceives,  at  least  to  a  minimal  degree,  conditions  4 
and  53  the  unconditional  positive  regard  of  the  therapist  for  him,  and  the 
empathic  understanding  of  the  therapist. 

Comment.  These  seem  to  be  the  necessary  conditions  of  therapy, 
though  other  elements  are  often  or  usually  present.  The  process  is  more 
likely  to  get  under  way  if  the  client  is  anxious,  rather  than  merely 
vulnerable.  Often  it  is  necessary  for  the  contact  or  relationship  to  be  of 
some  duration  before  the  therapeutic  process  begins.  Usually  the  em- 
pathic understanding  is  to  some  degree  expressed  verbally,  as  well  as 
experienced.  But  the  process  often  commences  with  only  these  minimal 
conditions,  and  it  is  hypothesized  that  it  never  commences  without  these 
conditions  being  met. 

The  point  which  is  most  likely  to  be  misunderstood  is  the  omission 
of  any  statement  that  the  therapist  communicates  his  empathic  under- 
standing and  his  unconditional  positive  regard  to  the  client.  Such  a  state- 
ment has  been  omitted  only  after  much  consideration,  for  these  reasons. 
It  is  not  enough  for  the  therapist  to  communicate,  since  the  communica- 
tion must  be  received,  as  pointed  out  in  condition  6,  to  be  effective. 
It  is  not  essential  that  the  therapist  intend  such  communication,  since 
often  it  is  by  some  casual  remark,  or  involuntary  facial  expression,  that 
the  communication  is  actually  achieved.  However,  if  one  wishes  to  stress 
the  communicative  aspect  which  is  certainly  a  vital  part  of  the  living 
experience,  then  condition  6  might  be  worded  in  this  fashion : 

6.  That  the  communication  to  the  client  of  the  therapist's  empathic 
understanding  and  unconditional  positive  regard  is,  at  least  to  a  minimal 
degree,  achieved. 

The  element  which  will  be  most  surprising  to  conventional  therapists 
is  that  the  same  conditions  are  regarded  as  sufficient  for  therapy,  regard- 
less of  the  particular  characteristics  of  the  client.  It  has  been  our  ex- 
perience to  date  that  although  the  therapeutic  relationship  is  used  dif- 

214  CARL    R.    ROGERS 

ferently  by  different  clients,  it  is  not  necessary  nor  helpful  to  manipulate 
the  relationship  in  specific  ways  for  specific  kinds  of  clients.  To  do  this 
damages,  it  seems  to  us,  the  most  helpful  and  significant  aspect  of  the 
experience,  that  it  is  a  genuine  relationship  between  two  persons,  each 
of  whom  is  endeavoring,  to  the  best  of  his  ability,  to  be  himself  in  the 

The  "growing  edge"  of  this  portion  of  the  theory  has  to  do  with 
point  3,  the  congruence  or  genuineness  of  the  therapist  in  the  relation- 
ship. This  means  that  the  therapist's  symbolization  of  his  own  ex- 
perience in  the  relationship  must  be  accurate,  if  therapy  is  to  be  most 
effective.  Thus  if  he  is  experiencing  threat  and  discomfort  in  the 
relationship,  and  is  aware  only  of  an  acceptance  and  understanding,  then 
he  is  not  congruent  in  the  relationship  and  therapy  will  suffer.  It  seems 
important  that  he  should  accurately  "be  himself3  in  the  relationship, 
whatever  the  self  of  that  moment  may  be. 

Should  he  also  express  or  communicate  to  the  client  the  accurate 
symbolization  of  his  own  experience?  The  answer  to  this  question  is  still 
in  an  uncertain  state.  At  present  we  would  say  that  such  feelings  should 
be  expressed,  if  the  therapist  finds  himself  persistently  focused  on  his  own 
feelings  rather  than  those  of  the  client,  thus  greatly  reducing  or  eliminat- 
ing any  experience  of  empathic  understanding,  or  if  he  finds  himself 
persistently  experiencing  some  feeling  other  than  unconditional  positive 
regard.  To  know  whether  this  answer  is  correct  demands  further  testing 
of  the  hypothesis  it  contains,  and  this  is  not  simple  since  the  courage 
to  do  this  is  often  lacking,  even  in  experienced  therapists.  When  the  thera- 
pist's real  feelings  are  of  this  order:  "I  find  myself  fearful  that  you  are 
slipping  into  a  psychosis,"  or  CCI  find  myself  frightened  because  you  are 
touching  on  feelings  I  have  never  been  able  to  resolve,"  then  it  is  difficult 
to  test  the  hypothesis,  for  it  is  very  difficult  for  the  therapist  to  express 
such  feelings. 

Another  question  which  arises  is  this :  is  it  the  congruence,  the  whole- 
ness, the  integration  of  the  therapist  in  the  relationship  which  is  im- 
portant, or  are  the  specific  attitudes  of  empathic  understanding  and  un- 

1This  paragraph  may  have  to  be  rewritten  if  a  recent  study  of  Klrtncr  [42] 
is  confirmed.  Kirtner  has  found,  in  a  group  of  26  cases  from  the  Counseling 
Center  at  the  University  of  Chicago,  that  there  are  sharp  differences  in  the  client's 
mode  of  approach  to  the  resolution  of  life  difficulties  and  that  these  differences 
are  related  to  success  in  therapy.  Briefly,  the  client  who  sees  his  problem  as  in- 
volving his  relationships,  and  who  feels  that  he  contributes  to  this  problem  and 
wants  to  change  it,  is  likely  to  be  successful.  The  client  who  externalizes  his 
problem  and  feels  little  self-responsibility  is  much  more  likely  to  be  a  failure. 
Thus  the  implication  is  that  different  conditions  of  therapy  may  be  necessary  to 
make  personality  change  possible  in  this  latter  group.  If  this  is  verified,  then  the 
theory  will  have  to  be  revised  accordingly. 

Therapy,  Personality,  and  Interpersonal  Relationships  215 

conditional  positive  regard  vital?  Again  the  final  answer  is  unknown,  but 
a  conservative  answer,  the  one  we  have  embodied  in  the  theory,  is  that 
for  therapy  to  occur  the  wholeness  of  the  therapist  in  the  relationship  is 
primary,  but  a  part  of  the  congruence  of  the  therapist  must  be  the  ex- 
perience of  unconditional  positive  regard  and  the  experience  of  empathic 

Another  point  worth  noting  is  that  the  stress  is  upon  the  experience 
in  the  relationship.  It  is  not  to  be  expected  that  the  therapist  is  a  com- 
pletely congruent  person  at  all  times.  Indeed  if  this  were  a  necessary 
condition  there  would  be  no  therapy.  But  it  is  enough  if  in  this  particular 
moment  of  this  immediate  relationship  with  this  specific  person  he  is 
completely  and  fully  himself,  with  his  experience  of  the  moment  being 
accurately  symbolized  and  integrated  into  the  picture  he  holds  of  himself. 
Thus  it  is  that  imperfect  human  beings  can  be  of  therapeutic  assistance 
to  other  imperfect  human  beings. 

The  greatest  flaw  in  the  statement  of  these  conditions  is  that  they 
are  stated  as  if  they  were  all-or-none  elements,  whereas  conditions  2  to  6 
all  exist  on  continua.  At  some  later  date  we  may  be  able  to  say  that 
the  therapist  must  be  genuine  or  congruent  to  such  and  such  a  degree 
in  the  relationship,  and  similarly  for  the  other  items.  At  the  present  we 
can  only  point  out  that  the  more  marked  the  presence  of  conditions  2  to 
6,  the  more  certain  it  is  that  the  process  of  therapy  will  get  under  way, 
and  the  greater  the  degree  of  reorganization  which  will  take  place.  This 
function  can  only  be  stated  qualitatively  at  the  present  time. 

Evidence.  Confirmatory  evidence,  particularly  of  item  5,  is  found 
in  the  studies  by  Fiedler  [19,  20]  and  Quinn  [52].  Fiedler's  study  showed 
that  experienced  therapists  of  different  orientations  created  relationships 
in  which  one  of  the  most  prominent  characteristics  was  the  ability  to 
understand  the  client's  communications  with  the  meaning  these  com- 
munications had  for  the  client.  Quinn  found  that  the  quality  of  therapist 
communication  was  of  crucial  significance  in  therapy.  These  studies  add 
weight  to  the  importance  of  empathic  understanding. 

Seeman  [75]  found  that  increase  in  the  counselor's  liking  for  the 
client  during  therapy  was  significantly  associated  with  therapeutic  success. 
Both  Seeman  and  Lipkin  [44]  found  that  clients  who  felt  themselves  to 
be  liked  by  the  therapist  tended  to  be  more  successful.  These  studies 
tend  to  confirm  condition  4  (unconditional  positive  regard)  and  condi- 
tion 6  (perception  of  this  by  the  client) . 

Though  clinical  experience  would  support  condition  2,  the  client's 
vulnerability  or  anxiety,  there  is  little  research  which  has  been  done  in 
terms  of  these  constructs.  The  study  by  Gallagher  [21]  indicates  that 
less  anxious  clients  tend  never  to  become  involved  in  therapy,  but  drop 

216  CARL  R.   ROGERS 

B.  The  Process  of  Therapy 

When  the  preceding  conditions  exist  and  continue,  a  process  is  set  in 
motion  which  has  these  characteristic  directions : 

1.  The   client  is   increasingly  free  in   expressing  his  feelings.,   through 
verbal  and/or  motor  channels. 

2.  His  expressed  feelings  increasingly  have  reference  to  the  self,  rather 
than  nonself . 

3.  He  increasingly  differentiates  and  discriminates  the  objects  of  his 
feelings  and  perceptions.,  including  his  environment,  other  persons,  his  self, 
his  experiences,  and  the  interrelationships  of  these.  He  becomes  less  in- 
tenslonal  and  more  extensional  in  his  perceptions,  or  to  put  it  in  other  terms, 
his  experiences  are  more  accurately  symbolized. 

4.  His  expressed  feelings  increasingly  have  reference  to  the  incongruity 
between  certain  of  his  experiences  and  his  concept  of  self. 

5.  He  comes  to  experience  in  awareness  the  threat  of  such  incongruence. 
a.  This  experience  of  threat  is  possible  only  because  of  the  continued 

unconditional  positive  regard  of  the  therapist,  which  is  extended  to 
incongruence  as  much  as  to  congruence,  to  anxiety  as  much  as  to 
absence  of  anxiety. 

6.  He  experiences  fully,  in  awareness,  feelings  which  have  in  the  past 
been  denied  to  awareness,  or  distorted  in  awareness. 

7.  His  concept  of  self  becomes  reorganized  to  assimilate  and  include 
these  experiences  which  have  previously  been  distorted  in  or  denied  to 

8.  As  this  reorganization  of  the  self-structure  continues,  his  concept  of 
self  becomes  increasingly  congruent  with  his  experience;  the  self  now  in- 
cluding experiences  which  previously  would  have  been  too  threatening  to  be 
in  awareness. 

a.  A  corollary  tendency  is  toward  fewer  perceptual  distortions  in  aware- 
ness, or  denials  to  awareness,  since  there  are  fewer  experiences  which 
can  be  threatening.  In  other  words,  defensiveness  is  decreased. 

9.  He  becomes  increasingly  able  to   experience,  without  a  feeling  of 
threat,  the  therapist's  unconditional  positive  regard. 

10.  He  increasingly  feels  an  unconditional  positive  self-regard. 

11.  He  increasingly  experiences  himself  as  the  locus  of  evaluation. 

12.  He  reacts  to  experience  less  in  terms  of  his  conditions  of  worth  and 
more  in  terms  of  an  organismlc  valuing  process. 

Comment.  It  cannot  be  stated  with  certainty  that  all  of  these  are 
necessary  elements  of  the  process,  though  they  are  all  characteristic.  Both 
from  the  point  of  view  of  experience,  and  the  logic  of  the  theory,  3,  6,  73 

Therapy,  Personality,  and  Interpersonal  Relationships  217 

8,  10,  12,  are  necessary  elements  in  the  process.  Item  5a  is  not  a  logical 
step  in  the  theory  but  is  put  in  as  an  explanatory  note. 

The  element  which  will  doubtless  be  most  puzzling  to  the  reader 
is  the  absence  of  explanatory  mechanisms.  It  may  be  well  to  restate  our 
scientific  purpose  in  terms  of  an  example.  //  one  strokes  a  piece  of  steel 
with  a  magnet,  and  if  one  places  the  piece  of  steel  so  that  it  can  rotate 
freely,  then  it  will  point  to  the  north.  This  statement  of  the  if-then 
variety  has  been  proved  thousands  of  times.  Why  does  it  happen?  There 
have  been  various  theoretical  answers,  and  one  would  hesitate  to  say, 
even  now,  that  we  know  with  certitude  why  this  occurs. 

In  the  same  way  I  have  been  saying  in  regard  to  therapy,  "If  these 
conditions  exist,  then  these  subsequent  events  will  occur."  Of  course  we 
have  speculations  as  to  why  this  relationship  appears  to  exist,  and  those 
speculations  will  be  increasingly  spelled  out  as  the  presentation  continues. 
Nevertheless  the  most  basic  element  of  our  theory  is  that  if  the  described 
conditions  exist,  then  the  process  of  therapy  occurs,  and  the  events 
which  are  called  outcomes  will  be  observed.  We  may  be  quite  wrong 
as  to  why  this  sequence  occurs.  I  believe  there  is  an  increasing  body  of 
evidence  to  show  that  it  does  occur. 

Evidence.  There  is  confirming  evidence  of  varying  degrees  of 
relevance  for  a  number  of  these  items  describing  the  therapeutic  process. 
Item  2  (increasing  self-reference)  is  supported  by  our  many  recorded 
therapeutic  cases,  but  has  not  been  reduced  to  a  statistical  finding. 
Stock's  study  [82]  supports  item  3,  indicating  that  client  self-referent 
expressions  become  more  objective,  less  strongly  emotional.  Mitchell  [47] 
shows  that  clients  become  more  extensional. 

Objective  clinical  evidence  supporting  items  4,  5,  and  6  is  provided 
in  the  form  of  recordings  from  a  case  by  Rogers  [67]. 

The  findings  of  Vargas  [85]  are  relevant  to  item  7,  indicating  the 
way  the  self  is  reorganized  in  terms  of  emergent  new  self-perceptions. 
Hogan  [36]  and  Haigh  [29]  have  studied  the  decrease  in  defensiveness 
during  the  process,  as  described  in  item  8a,  their  findings  being  con- 
firmatory. The  increased  congruence  of  self  and  experience  is  supported 
in  an  exhaustive  single  case  investigation  by  Rogers  [67].  That  such 
congruence  is  associated  with  lack  of  defensiveness  is  found  by  Chodor- 
koflf  [10]. 

Item  10,  the  increase  in  the  client's  positive  self-regard,  is  well 
attested  by  the  studies  of  Snyder  [79],  Seeman  [76],  Raimy  [55],  Stock 
[82],  Strom  [83],  Sheerer  [78],  Lipkin  [44].  The  client's  trend  toward 
experiencing  himself  as  the  locus  of  evaluation  is  most  clearly  shown 
by  Raskin's  research  [56],  but  this  is  supported  by  evidence  from  Sheerer 
[78],  Lipkin  [44],  Kessler  [41]. 

218  CARL   R.    ROGERS 

C.  Outcomes  in  Personality  and  Behavior 

There  is  no  clear  distinction  between  process  and  outcome.  Items  of 
process  are  simply  differentiated  aspects  of  outcome.  Hence  the  statements 
which  follow  could  have  been  included  under  process.  For  reasons  of  con- 
venience in  understanding,  there  have  been  grouped  here  those  changes 
which  are  customarily  associated  with  the  terms  outcomes,  or  results,  or  are 
observed  outside  of  the  therapeutic  relationship.  These  are  the  changes 
which  are  hypothesized  as  being  relatively  permanent : 

1.  The   client  is  more   congruent,  more   open   to   his  experience,   less 

2.  He  is  consequently  more  realistic,  objective,  extensional  in  his  per- 

3.  He  is  consequently  more  effective  in  problem  solving, 

4.  His    psychological    adjustment    is    improved,    being    closer    to    the 

a.  This  is  owing  to,  and  is  a  continuation  of,  the  changes  in  self-structure 
described  in  B7  and  BB. 

5.  As  a  result  of  the  increased  congruence  of  self  and  experience  (C4 
above)  his  vulnerability  to  threat  is  reduced. 

6.  As  a  consequence  of  C2  above,  his  perception  of  his  ideal  self  is 
more  realistic,  more  achievable. 

7.  As  a  consequence  of  the  changes  in  C4  and  C5  his  self  is  more  con- 
gruent with  his  ideal  self. 

8.  As  a  consequence  of  the  increased  congruence  of  self  and  ideal  self 
(C6)  and  the  greater  congruence  of  self  and  experience,  tension  of  all  types 
is  reduced — physiological  tension,  psychological  tension,  and  the  specific 
type  of  psychological  tension  defined  as  anxiety. 

9.  He  has  an  increased  degree  of  positive  self-regard. 

10.  He  perceives  the  locus  of  evaluation  and  the  locus  of  choice  as 
residing  within  himself. 

a.  As  a  consequence  of  C9  and  CIO  he  feels  more  confident  and  more 

b.  As  a  consequence  of  Cl  and  CIO,  his  values  are  determined  by  an 
organismic  valuing  process. 

11.  As  a  consequence  of  C13  and  C2,  he  perceives  others  more  realisti- 
cally and  accurately. 

12.  He  experiences  more  acceptance  of  others,  as  a  consequence  of  less 
need  for  distortion  of  his  perceptions  of  them. 

13.  His  behavior  changes  in  various  ways. 

a.  Since  the  proportion  of  experience  assimilated  into  the  self -structure 
is  increased,  the  proportion  of  behaviors  which  can  be  "owned"  as 
belonging  to  the  self  is  increased. 

Therapy,  Personality,  and  Interpersonal  Relationships  219 

b.  Conversely,  the  proportion  of  behaviors  which  are  disowned  as  self- 
experiences,  felt  to  be  "not  myself/'  is  decreased. 

c.  Hence  his  behavior  is  perceived  as  being  more  within  his  control. 

14.  His  behavior  is  perceived  by  others  as  more  socialized,  more  mature. 

15.  As  a  consequence  of  Cl,  2,  3,  his  behavior  is  more  creative,  more 
uniquely  adaptive  to  each  new  situation,  and  each  new  problem3  more  fully 
expressive  of  his  own  purposes  and  values. 

Comment.  The  statement  in  part  C  which  is  essential  is  statement  Cl. 
Items  2  through  15  are  actually  a  more  explicit  spelling  out  of  the 
theoretical  implications  of  statement  1.  The  only  reason  for  including 
them  is  that  though  such  implications  follow  readily  enough  from  the 
logic  of  the  theory,  they  are  often  not  perceived  unless  they  are  pointed 

Evidence.  There  is  much  confirmatory  and  some  ambiguous  or  non- 
confirming  evidence  of  the  theoretical  statement  of  outcomes.  Grummon 
and  John  [28]  find  a  decrease  in  defensiveness,  basing  judgements  on 
the  TAT.  Hogan  [36]  and  Haigh  [29]  also  supply  some  scanty  evidence 
on  this  point.  As  to  the  greater  extensionality  of  perceptions  (item  2), 
Jonietz  [38]  finds  that  therapy  produces  changes  in  perceptions  and 
Mitchell  [47]  finds  these  changes  to  be  in  the  direction  of  extensionality. 

Item  4,  stating  that  adjustment  is  improved,  is  supported  by  evidence 
based  upon  TAT,  Rorschach,  counselor  rating,  and  other  indexes,  in  the 
studies  of  Dymond  [15,  16],  Grummon  and  John  [28],  Haimowitz  [30], 
Muench  [49],  Mosak  [48],  Cowen  and  Combs  [13].  Carr  [8],  however, 
found  no  evidence  of  change  in  the  Rorschach  in  nine  cases. 

Rudikoff  [73]  found  that  the  self-ideal  becomes  more  achievable, 
as  stated  in  item  6.  The  increased  congruence  of  self  and  ideal  has  been 
confirmed  by  Butler  and  Haigh  [7],  Hartley  [33],  and  its  significance  for 
adjustment  supported  by  Hanlon,  Hofstaetter,  and  O'Connor  (32). 

The  decrease  in  physiological  tension  over  therapy  is  attested  by 
the  studies  of  Thetford  [84]  and  Anderson  [1].  The  reduction  in  psy- 
chological tension  as  evidenced  by  the  Discomfort-Relief  Quotient  has 
been  confirmed  by  many  investigators :  Assum  and  Levy  [4] ,  Gofer  and 
Chance  [12],  Kaufman  and  Raimy  [39],  N.  Rogers  [72],  Zimmerman 

The  increase  in  positive  self-regard  is  well  attested,  as  indicated  in 
IB,  Evidence.  The  shift  in  the  locus  of  evaluation  and  choice  is  supported 
in  the  evidence  provided  by  Raskin  [56]  and  Sheerer  [78].  Rudikoff  [73] 
presents  evidence  which  suggests  that  others  may  be  perceived  with 
greater  realism.  Sheerer  [78]  and  Stock  [82]  and  Rudikoff  [73]  show 
that  others  are  perceived  in  a  more  acceptant  fashion  as  postulated 
in  item  11.  Gordon  and  Cartwright  [25]  provide  evidence  which  is 

220  CARL   R.    ROGERS 

complex  but  In  general  nonconfirming  on  this  point.  M.  Haimowitz 
[30]  also  has  findings  which  seem  to  indicate  that  nonacceptance  of 
minority  groups  may  be  more  openly  expressed. 

The  behavior  changes  specified  in  items  13  and  14  find  support  in 
the  Rogers  study  [68]  showing  that  in  improved  cases  both  the  client 
and  his  friends  observe  greater  maturity  in  his  behavior.  Hoffman  [35] 
finds  that  the  behavior  the  client  describes  in  the  interviews  becomes 
more  mature.  Jonietz's  study  of  [38]  of  perception  of  ink  blots  might  lend 
some  support  to  the  postulate  of  item  15. 

Comments  on  the  theory  of  therapy.  It  is  to  be  noted  that  this 
theory  of  therapy  involves,  basically,  no  intervening  variables.  The  condi- 
tions of  therapy,  given  in  A,  are  all  operationally  definable,  and  some 
have  already  been  given  rather  crude  operational  definitions  in  research 
already  conducted.  The  theory  states  that  if  A  exists,  then  B  and  C  will 
follow.  B  and  C  are  measurable  events,  predicted  by  A. 

It  should  also  be  pointed  out  that  the  logic  of  the  theory  is  such  that : 
if  A,  then  B;  if  A,  then  B  and  C;  if  A,  then  C  (omitting  conisderation  of 
5),  if  B.,  then  C  (omitting  consideration  of  A) . 

Specification  of  functional  relationships.  At  this  point,  the  functional 
relationships  can  only  be  stated  in  general  and  qualitative  form.  The 
greater  the  degree  of  the  conditions  specified  in  A}  the  more  marked 
or  more  extensive  will  be  the  process  changes  in  B,  and  the  greater  or 
more  extensive  the  outcome  changes  specified  in  C.  Putting  this  in  more 
general  terms,  the  greater  the  degree  of  anxiety  in  the  client,  congruence 
in  the  therapist  in  the  relationship,  acceptance  and  empathy  experienced 
by  the  therapist,  and  recognition  by  the  client  of  these  elements,  the 
deeper  will  be  the  process  of  therapy,  and  the  greater  the  extent  of 
personality  and  behavioral  change.  To  revert  now  to  the  theoretical 
logic,  all  we  can  say  at  present  is  that 

B  =  (f)A        C=(f)A        B  +  C  =  (f)A        C  =  (f)B 

Obviously  there  are  many  functional  interrelationships  not  yet 
specified  by  the  theory.  For  example,  if  anxiety  is  high,  is  congruence  on 
the  part  of  the  therapist  less  necessary?  There  is  much  work  to  be  done 
in  investigating  the  functional  relationships  more  fully. 

D.  Some  Conclusions  Regarding  the  Nature  of  the  Individual 

From  the  theory  of  therapy  as  stated  above,  certain  conclusions  are 
implicit  regarding  the  nature  of  man.  To  make  them  explicit  involves  little 
more  than  looking  at  the  same  hypotheses  from  a  somewhat  different 
vantage  point.  It  is  well  to  state  them  explicitly,  however,  since  they  con- 
stitute an  important  explanatory  link  of  a  kind  which  gives  this  theory  what- 

Therapy,  Personality,  and  Interpersonal  Relationships  221 

ever  uniqueness  it  may  possess.  They  also  constitute  the  impelling  reason  for 
developing  a  theory  of  personality.  If  the  individual  is  what  he  is  revealed 
to  be  in  therapy,  then  what  theory  would  account  for  such  an  individual? 
We  present  these  conclusions  about  the  characteristics  of  the  human 
organism : 

1.  The  individual  possesses  the  capacity  to  experience  in  awareness  the 
factors  in  his  psychological  maladjustment,  namely,  the  incongruences  be- 
tween his  self-concept  and  the  totality  of  his  experience. 

2.  The  individual  possesses  the  capacity  and  has  the  tendency  to  re- 
organize his  self-concept  in  such  a  way  as  to  make  it  more  congruent  with 
the  totality  of  his  experience,  thus  moving  himself  away  from  a  state  of 
psychological  maladjustment,  and  toward  a  state  of  psychological  adjust- 

3.  These  capacities  and  this  tendency,  when  latent  rather  than  evident, 
will  be  released  in  any  interpersonal  relationship  in  which  the  other  person 
is  congruent  in  the  relationship,  experiences  unconditional  positive  regard 
toward,  and  em  pat  hie  understanding  of  the  individual,  and  achieves  some 
communication  of  these  attitudes  to  the  individual.  (These  are,  of  course, 
the  characteristics  already  given  under  1,43,  4,  5,  6.) 

It  is  this  tendency  which,  in  the  following  theory  of  personality,  is 
elaborated  into  the  tendency  toward  actualization. 

I  believe  it  is  obvious  that  the  basic  capacity  which  is  hypothesized 
is  of  very  decided  importance  in  its  psychological  and  philosophical  im- 
plications. It  means  that  psychotherapy  is  the  releasing  of  an  already 
existing  capacity  in  a  potentially  competent  individual,  not  the  expert 
manipulation  of  a  more  or  less  passive  personality.2  Philosophically  it 
means  that  the  individual  has  the  capacity  to  guide,  regulate,  and  control 
himself,  providing  only  that  certain  definable  conditions  exist.  Only  in 
the  absence  of  these  conditions,  and  not  in  any  basic  sense,  is  it  necessary 
to  provide  external  control  and  regulation  of  the  individual. 


In  endeavoring  to  order  our  perceptions  of  the  individual  as  he 
appears  in  therapy,  a  theory  of  the  development  of  personality,  and  of  the 
dynamics  of  behavior,  has  been  constructed.  It  may  "be  well  to  repeat 
the  warning  previously  given,  and  to  note  that  the  initial  propositions 

2  In  order  to  correct  a  common  misapprehension  it  should  be  stated  that  this 
tentative  conclusion  in  regard  to  human  capacity  grew  out  of  continuing  work 
with  clients  in  therapy.  It  was  not  an  assumption  or  bias  with  which  we  started 
our  therapeutic  endeavors.  A  brief  personal  account  of  the  way  in  which  this 
conclusion  was  forced  upon  me  is  contained  in  an  autobiographical  paper  [69], 

222  CARL    R.    ROGERS 

of  this  theory  are  those  which  are  furthest  from  the  matrix  of  our  ex- 
perience and  hence  are  most  suspect.  As  one  reads  on,  the  propositions 
become  steadily  closer  to  the  experience  of  therapy.  As  before,  the  defined 
terms  and  constructs  are  italicized,  and  are  to  be  understood  as  previously 

A.  Postulated  Characteristics  of  the  Human  Infant 

It  is  postulated  that  the  individual,  during  the  period  of  infancy,  has  at 
least  these  attributes. 

1.  He  perceives  his  experience  as  reality.  His  experience  is  his  reality. 
a.  As  a  consequence  he  has  greater  potential  awareness  of  what  reality 
is  for  him  than  does  anyone  else,  since  no  one  else  can  completely 
assume  his  internal  frame  of  referenced-     '     ^ 
*>  2.  He  has  an  inherent  tendency  toward  actualizing  his  organism. 

3.  He  interacts  with  his  reality  in  terms  of  his  basic  actualizing  tendency. 
Thus  his  behavior  is  the  goal-directed  attempt  of  the  organism  to  satisfy 
the  experienced  needs  for  actualization  in  the  reality  as  perceived. 
'?.  4.  In  this  interaction  he  behaves  as  an  organized  whole,  as  a  gestalt. 

5.  He  engages  in  an  organismic  valuing  process,  valuing  experience  with 
reference  to  the  actualizing  tendency  as  a  criterion.  Experiences  which  are 
perceived  as  maintaining  or  enhancing  the  organism are"valued  positively. 
Those  which  are  perceived  as  negating  such  maintenance  or  enhancement 
are  valued  negatively. 

6.  He  behaves  with  adience  toward  positively  valued  experiences  and 
with  avoidance  toward  those  negatively  valued. 

Comment.  In  this  view  as  formally  stated,  the  human  infant  is  seen 
as  having  an  inherent  motivational  system  (which  he  shares  in  common 
with  all  living  things)  and  a  regulatory  system  (the  valuing  process) 
which  by  its  "feedback"  keeps  the  organism  "on  the  beam33  of  satis- 
fying his  motivational  needs.  He  lives  in  an  environment  which  for  the- 
oretical purposes  may  be  said  to  exist  only  in  him,  or  to  be  of  his  own 

This  last  point  seems  difficult  for  some  people  to  comprehend.  It  is 
the  perception  of  the  environment  which  constitutes  the  environment, 
regardless  as  to  how  this  relates  to  some  "real33  reality  which  we  may 
philosophically  postulate.  The  infant  may  be  picked  up  by  a  friendly, 
affectionate  person.  If  his  perception  of  the  situation  is  that  this  is  a 
strange  and  frightening  experience,  it  is  this  perception,  not  the  "reality" 
or  the  "stimulus"  which  will  regulate  his  behavior.  To  be  sure,  the  rela- 
tionship with  the  environment  is  a  transactional  one,  and  if  his  con- 
tinuing experience  contradicts  his  initial  perception,  then  in  time  his 

Therapy,  Personality,  and  Interpersonal  Relationships  223 

perception  will  change.  But  the  effective  reality  which  influences  behavior 
is  at  all  times  the  perceived  reality.  We  can  operate  theoretically  from 
this  base  without  having  to  resolve  the  difficult  question  of  what  "really" 
constitutes  reality. 

Another  comment  which  may  be  in  order  is  that  no  attempt  has 
been  made  to  supply  a  complete  catalogue  of  the  equipment  with  which 
the  infant  faces  the  world.  Whether  he  possesses  instincts,  or  an  in- 
nate sucking  reflex,  or  an  innate  need  for  affection,  are  interesting 
questions  to  pursue,  but  the  answers  seem  peripheral  rather  than  essential 
to  a  theory  of  personality. 

B.  The  Development  of  the  Self 

1.  In  line  with  the  tendency  toward  differentiation  which  is  a  part  of 
the  actualizing  tendency,  a  portion  of  the  individual's  experience  becomes 
differentiated  and  symbolized  in  an  .awareness  of  being,  awareness  of  func- 
tioning. Such  awareness  may  be  described  as  self-experience. 

2.  This  representation  in  awareness  of  being  and  functioning,  becomes 
elaborated,  through  interaction  with  the  environment,  particularly  the  en- 
vironment composed  of  significant  others,  into  a  concept  of  self,  a  perceptual 
object  in  his  experiential  field. 

Comment.  These  are  the  logical  first  steps  in  the  development  of  the 
self.  It  is  by  no  means  the  way  the  construct  developed  in  our  own 
thinking,  as  has  been  indicated  in  the  section  of  definitions.  (A  digression 
on  the  case  history  of  a  construct,  p.  200. ) 

C.  The  Need  for  Positive  Regard 

1.  As  the  awareness  of  self  emerges,  the  individual  develops  a  need  for 
positive  regard.  This  need  is  universal  in  human  beings,  and  in  the  in- 
dividual, is  pervasive  and  persistent.  Whether  it  is  an  inherent  or  learned 
need  is  irrelevant  to  the  theory.  Standal  [80],  who  formulated  the  concept, 
regards  it  as  the  latter. 

a.  The  satisfaction  of  this  need  is  necessarily  based  upon  inferences  re- 
garding the  experiential  field  of  another. 

( 1 )    Consequently  it  is  often  ambiguous. 

b.  It  is  associated  with  a/  very  wide  range  of  the  individual's  experiences. 

c.  It  is  reciprocal,  in  that  when  an  individual  discriminates  himself  as 
satisfying  another's  need  for  positive  regard,  he  necessarily  experiences 
satisfaction  of  his  own  need  for  positive  regard. 

( 1 )    Hence  it  is  rewarding  both  to  satisfy  this  need  in  another,  and  to 
experience  the  satisfaction  of  one's  own  need  by  another. 

224  CARL   R.    ROGERS 

d.  It  is  potent,,  in  that  the  positive  regard  of  any  social  other  is  com- 
municated to  the  total  regard  complex  which  the  individual  associates 
with  that  social  other. 

(1)  Consequently  the  expression  of  positive  regard  by  a  significant 
social  other  can  become  more  compelling  than  the  organismic 
valuing  process,  and  the  individual  becomes  more  adient  to 
the  positive  regard  of  such  others  than  toward  experiences  which 
are  of  positive  value  in  actualizing  the  organism. 

D.  The  Development  of  the  Need  for  Self-regard 

1.  The  positive  regard  satisfactions  or  frustrations  associated  with  any 
particular  self-experience  or  group  of  self-experiences  come  to  be  experienced 
by  the  individual  independently  of  positive  regard  transactions  with  social 
others.  Positive  regard  experienced  in  this  fashion  is  termed  self-regard. 

2.  A  need  for  self-regard  develops  as  a  learned  need  developing  out  of 
the  association  of  self-experiences  with  the  satisfaction  or  frustration  of  the 
need  for  positive  regard.  —»«•-• 

3.  The  individual  thus  comes  to  experience  positive  regard  or  loss  of 
positive   regard  independently  of  transactions  with  any  social  other.   He 
becomes  in  a  sense  his  own  significant  social  other. 

4.  Like  positive  regard,  self-regard  which  is  experienced  in  relation  to 
any  particular  self-experience  or  group  of  self-experiences,  is  communicated 
to  the  total  self-regard  complex. 

E.  The  Development  of  Conditions  of  Worth 

1.  When  self -experiences  of  the  individual   are  discriminated  by  sig- 
nificant others  as  being  more  or  less  worthy  of  positive  regard,  then  self- 
regard  becomes  similarly  selective. 

2.  When  a  self-experience  is  avoided   (or  sought)    solely  because  it  is 
less  (or  more)  worthy  of  self -regard,  the  individual  is  said  to  have  acquired 
a  condition  of  worth. 

3.  If  an  individual  should  experience  only  unconditional  positive  regard, 
then  no  conditions  of  worth  would  develop,  self-regard  would  be  uncondi- 
tional, the  needs  for  positive  regard  and  self-regard  would  never  be  at 
variance  with  jgrfamsmic^n^  the  individual  would  continue  to 
be  psychologically  adjusted.,  and  would  be  fully  functioning.  This  chain  of 
events  is  hypothetically  possible,  and  hence  important  theoretically,  though 
it  does  not  appear  to  occur  in  actuality. 

Comment.  This  is  an  important  sequence  in  personality  development, 
stated  more  fully  by  Standal  [80].  It  may  help  to  restate  the  sequence  in 
informal,  illustrative,  and  much  less  exact  terms. 

Therapy,  Personality,  and  Interpersonal  Relationships  225 

The  infant  learns  to  need  love.  Love  is  very  satisfying,  but  to  know 
whether  he  is  receiving  it  or  not  he  must  observe  his  mother's  face, 
gestures,  and  other  ambiguous  signs.  He  develops  a  total  gestalt  as  to  the 
way  he  is  regarded  by  his  mother  and  each  new  experience  of  love  or 
rejection  tends  to  alter  the  whole  gestalt.  Consequently  each  behavior 
on  his  mother's  part  such  as  a  specific  disapproval  of  a  specific  behavior 
tends  to  be  experienced  as  disapproval  in  general.  So  important  is  this 
to  the  infant  that  he  comes  to  be  guided  in  his  behavior  not  by  the  degree 
to  which  an  experience  maintains  or  enhances  the  organism,  but  by  the 
likelihood  of  receiving  maternal  love. 

Soon  he  learns  to  view  himself  in  much  the  same  way,  liking  or 
disliking  himself  as  a  total  configuration.  He  tends,  quite  independently 
of  his  mother  or  atib^i^_to_yiew  himseli  andjns  behavior  in  the  same 
way  they  have.  This  means  that  some  behaviors  are  regarded  positively 
which  are  not  actually  experienced  organismically  as  satisfying.  Other 
behaviors  are  regarded  negatively  which  are  not  actually  experienced 
as  unsatisfying.  It  is  when  he  behaves  in  accordance  with  these  intro- 
jected  values  that  he  may  be  said  to  have  acquired  conditions_of_worth. 
He  cannot  regard  himself  positively,  as  having  worth,  unless  he  lives  in 
terms  of  these  conditions.  He  now  reacts  with  adience  or  avoidance 
toward  certain  behaviors  solely  because  of  these  introjected  conditions 
of  self-regard,  quite  without  reference  to  the  organismic  consequences 
of  these  behaviors.  This  is  what  is  meant  by  living  in  terms  of  intro- 
jected values  (the  phrase  formerly  used)  or  conditions  of  worth. 

It  is  not  theoretically  necessary  that  such  a  sequence  develop.  If  the 
infant  always  felt  prized,  if  his  own  feelings  were  always  accepted  even 
though  some  behaviors  were  inhibited,  then  no  conditions  of  worth 
would  develop.  This  could  at  least  theoretically  be  achieved  if  the 
parental  attitude  was  genuinely  of  this  sort:  "I  can  understand  how  satis- 
fying it  feels  to  you  to  hit  your  baby  brother  (or  to  defecate  when  and 
where  you  please,  or  to  destroy  things)  and  I  love  you  and  am  quite 
willing  for  you  to  have  those  feelings.  But  I  am  quite  willing  for  me  to 
have  my  feelings,  too,  and  I  feel  very  distressed  when  your  brother  is 
hurt,  (or  annoyed  or  sad  at  other  behaviors)  and  so  I  do  not  let  you  hit 
him.  Both  your  feelings  and  my  feelings  are  important,  and  each  of  us 
can  freely  have  his  own."  If  the  child  were  thus  able  to  retain  his  own 
organismic  evaluation  of  each  experience,  then  his  life  would 'become 
a  balancing  of  these  satisfactions.  Schematically  he  might  feel,  "I  enjoy 
hitting  baby  brother.  It  feels  good.  I  do  not  enjoy  mother's  distress.  That 
feels  dissatisfying  to  me.  I  enjoy  pleasing  her/'  Thus  his  behavior  would 
sometimes  involve  the  satisfaction  of  hitting  his  brother,  sometimes  the 
satisfaction  of  pleasing  mother.  But  he  would  never  have  to  disown  the 

226  CARL   R.    ROGERS 

feelings  of  satisfaction  or  dissatisfaction  which  he  experienced  in  this 
differential  way. 

F.  The  Development  of  Incongruence  between  Self  and  Experience 

1.  Because  of  the  need  for  self-regard,  the  individual  perceives  his  ex- 
perience selectively,  in  terms  of  the  conditions  of  worth  which  have  come  to 
exist  in  him. 

a.  Experiences  which  are  in  accord  with  his  conditions  of  worth  are  per- 
ceived and  symbolized  accurately  in  awareness. 

b.  Experiences  which  run  contrary  to  the  conditions  of  worth  are  per- 
ceived selectively  and  distortedly  as  if  in  accord  with  the  conditions  of 
worth,  or  are  in  part  or  whole,  denied  to  awareness. 

2.  Consequently  some  experiences  now  occur  in  the  organism  which  are 
not  recognized  as  self-experiences,  are  not  accurately  symbolized,  and  are  not 
organized  into  the  self-structure  in  accurately  symbolized  form. 

3.  Thus  from  the  time  of  the  first  selective  perception  in  terms  of  condi- 
tions of  worth,  the  states  of  Incongruence  between  self  and  experience,  of 
psychological  maladjustment  and  of  vulnerability,  exist  to  some  degree. 

Comment.  It  is  thus  because  of  the  distorted  perceptions  arising  from 
the  conditions  of  worth  that  the  individual  departs  from  the  integration 
which  characterizes  his  infant  state.  From  this  point  on  his  concept 
of  self  includes  distorted  perceptions  which  do  not  accurately  represent 
his  experience,  and  his  experience  includes  elements  which  are  not  in- 
cluded in  the  picture  he  has  of  himself.  Thus  he  can  no  longer  live  as  a 
unified  whole  person,  but  various  part  functions  now  become  char- 
acteristic. Certain  experiences  tend  to  threaten  the  self.  To  maintain 
the  self-structure  defensive  reactions  are  necessary.  Behavior  is  regulated 
at  times  by  the  self  and  at  times  by  those  aspects  of  the  organism's 
experience  which  are  not  included  in  the  self.  The  personality  is  hence- 
forth divided,  with  the  tensions  and  inadequate  functioning  which  ac- 
company such  lack  of  unity. 

This,  as  we  see  it,  is  the  basic  estrangement  in  man.  He  has  not  been 
true  to  himself,  to  his  own  natural  organismic  valuing  of  experience,  but 
for  the  sake  of  preserving  the  positive  regard  of  others  has  now  come 
to  f alsify^jpaie^pf  the  values  he  experiences  and  to  perceive  them  only 
in  terms  based  upon  ITfeff^value  "to  others.  Yet  this  has  not  been  a  con- 
scious choice,  but  a  natural — and  tragic — development  in  infancy.  The 
path  of  development  toward  psychological  maturity,  the  path  of  therapy, 
is  the  undoing  of  this  estrangement  in  man's  functioning,  the  dissolving 
°f  conditions  of  worth,  the  achievement  of  a  self  which  is  congruent 

Therapy,  Personality,,  and  Interpersonal  Relationships  227 

with  experience,  and  the  restoration  of  a  unified  organismic ,  valuing 
process  as  the  regulator  of  behavior.  "  ~~  "  ~~~*"~  ^ 

G.  The  Development  of  Discrepancies  in  Behavior 

1.  As  a  consequence  of  the  incongruence  between  self  and  experience 
described  in  F9  a  similar  incongruence  arises  in  the  behavior  of  the  in- 

a.  Some  behaviors  are  consistent  with  the  self-concept  and  maintain  and 
actualize  and  enhance  it. 

(1)    Such  behaviors  are  accurately  symbolized  in  awareness. 

b.  Some  behaviors  maintain,  enhance,  and  actualize  those  aspects  of  the 
experience  of  the  organism  which  are  not  assimilated  into  the  self- 

(1)   These  behaviors  are  either  unrecognized  as  self-experiences  or 
rceive^  z*n  distorted  or  selective  fashion  in  such  a  way  as  to  be 
congruent  with  the  self. 

H.  The  Experience  of  Threat  and  the  Process  of  Defense 

1.  As  the  organism  continues  to  experience,  an  experience  which  is  in- 
congruent  with  the  self-structure  (and  its  incorporated  conditions  of  worth] 
is  subceived  as  threatening. 

2.  The  essential  nature  of  the  threat  is  that  if  the  experience  were 
accurately  symbolized  in  awareness,  the  self-concept  would  no  longer  be  a 
consistent  gestalt,  the  conditions  of  worth  would  be  violated,  and  the  need 
for  self-regard  would  be  frustrated.  A  state  of  anxiety  would  exist. 

3.  The  process  of  defense  is  the  reaction  which  prevents  these  events 
from  occurring. 

a.  This  process  consists  of  the  selective  perception  or  distortion  of  the 

^         experience  and/or  the  denial  to  awareness  of  the  experience  or  some 

portion  thereof,  thus  keeping  the  total  perception  of  the  experience 

consistent  with  the  individual's  self -structure,  and  consistent  with  his 

conditions  of  worth. 

{J  4.  The  general  consequences  of  the  process  of  defense,  aside  from  its 
preservation  of  the  above  consistencies,  are  a  rigidity  of  perception,  due  to 
the  necessity  of  distorting  perceptions,  an  inaccurate  perception  of  reality, 
due  to  distortion  and  omission  of  data,  and  intensionality. 

Comment.  Section  G  describes  the  psychological  basis  for  what  are 
usually  thought  of  as  neuroti^^beliaviors,  and  Section  H  describes  the 
mechanisms  of  these  behaviors.  From  our  point  of  view  it  appears  more 
f undajnental  to  think  of  defensive  behaviors  ( described  in  these  two  sec- 
tions) and  disorganized  behaviors  (described  below).  Thus  the  de- 

228  CARL   R.    ROGERS 

fensive  behaviors  include  not  only  the  behaviors  customarily  regarded  as 
neurotic — rationalization,  compensation,  fantasy,  projection,  compul- 
sions, phobias,  and  the  like — but  also  some  of  the  behaviors  customarily 
regarded  as  psychotic,  notably  paranoid  behaviors  and  perhaps  catatonic 
^states.  /The  disorganized  category  includes  many  of  the  "irrational"  and 
"acute"  psychotic  behaviors,  as  will  be  explained  below.  This  seems  to  be 
a  more  fundamental  classification  than  those  usually  employed,  and 
perhaps  more  fruitful  in  considering  treatment.  It  also  avoids  any  con- 
cept of  neurosis  and  psychosis  as  entities  in  themselves,  which  we  believe 
has  been  an  unfortunate  and  misleading  conception. 

Let  us  consider  for  a  moment  the  general  range  of  the  defensive 
behaviors  from  the  simplest  variety,  common  to  all  of  us,  to  the  more 
extreme  and  crippling  varieties.  Take  first  of  all,  rationalization.  ("I 
didn't  really  make  that  mistake.  It  was  this  way.  .  .  .  ")  Such  excuses 
involve  a  perception  of  behavior  distorted  in  such  a  way  as  to  make  it 
congruent  with  our  concept  of  self  (as  a  person  who  doesn't  make 
mistakes).  Fantasy  is  another  example.  ("I  am  a  beautiful  princess,  and 
all  the  men  adore  me.")  Because  the  actual  experience  is  threatening  to 
the  concept  of  self  (as  an  adequate  person,  in  this  example),  this  ex- 
perience is  denied,  and  a  new  symbolic  world  is  created  which  enhances 
the  self,  but  completely  avoids  any  recognition  of  the  actual  experience. 
Where  the  incongruent  experience  is  a  strong  need,  the  organism 
actualizes  itself  by  finding  a  way  of  expressing  this  need,  but  it  is  per- 
ceived in  a  way  which  is  consistent  with  the  self.  Thus  an  individual 
whose  self-concept  involves  no  "bad"  sexual  thoughts  may  feel  or  ex- 
press the  thought  "I  am  pure,  but  you  are  trying  to  make  me  think 
filthy  thoughts."  This  would  be  thought  of  as  projection  or  as  a  paranoid 
idea.  It  involves  the  expression  of  the  organism's  need  for  sexual  satis- 
factions, but  it  is  expressed  in  such  a  fashion  that  this  need  may  be 
denied  to  awareness  and  the  behavior  perceived  as  consistent  with  the 
self.  Such  examples  could  be  continued,  but  perhaps  the  point  is  clear 
that  the  incongruence  between  self  and  experience  is  handled  by  the 
distorted  perception  of  experience  or  behavior,  or  by  the  denial  of  ex- 
perience in  awareness  (behavior  is  rarely  denied,  though  this  is  possible), 
or  by  some  combination  of  distortion  and  denial. 

7.  The  Process  of  Breakdown  and  Disorganization 

Up  to  this  point  the  theory  of  personality  which  has  been  formulated 
applies  to  every  individual  in  a  lesser  or  greater  degree.  In  this  and  the 
following  section  certain  processes  are  described  which  occur  _pnly  when 
certain  specified,  conditions  are  present. 

l~r  If  the  individual  has  a  large  or  significant  degree  of  mcongruencel 
between  self  and  experience  and  if  a  significant  experience  demonstrating  [ 

Therapy.,  Personality.,  and  Interpersonal  Relationships  229 

this  incongruence  occurs  suddenly,  or  with  a  high  degree  of  obviousness, 

then  the  organism's  process  of  defense  is  unable  to  operate  successfully. 

2.  As  a  result  anxiety  is  experienced  as  the  incongruence  is  subceived. 
The  degree  of  anxiety  is  dependent  upon  the  extent  of  the  self -structure 
which  is  threatened. 

3.  The  process  of  defense  being  unsuccessful,  the  experience  is  accurately 
symbolized  in  awareness,  and  the  gestalt  of  the  self-structure  is  broken  by 
this  experience  of  the  incongruence  in  awareness.  A  state  of  disorganization 

4.  In  such  a  state  of  disorganization  the  organism  behaves  at  times  in 
ways  which  are  openly  consistent  with  experiences  which  have  hitherto 
been  distorted  or  denied  to  awareness.  At  other  times  the  self  may  tem- 
porarily regain  regnancy,  and  the  organism  may  behave  in  ways  consistent 
with  it.  Thus  in  such  a  state  of  disorganization,  the  tension  between  the 
concept  of  self  (with  its  included  distorted  perceptions)  and  the  experiences 
which  are  not  accurately  symbolized  or  included  in  the  concept  of  self, 
is  expressed  in  a  confused^^grwicy,  first  one  and  jhen^the  other  supplying 
the  "feedback"  by  which  the  organism  regulates  behavior. 

Comment.  This  section,  as  will  be  evident  from  its  less  exact  for- 
mulation, is  new,  tentative,  and  needs  much  more  consideration.  Its 
meaning  can  be  illuminated  by  various  examples. 

Statements  1  and  2  above  may  be  illustrated  by  anxiety-producing 
experiences  in  therapy,  or  by  acute  psychotic  breakdowns.  In  the  free- 
dom of  therapy,  as  the  individual  expresses  more  and  more  of  himself, 
he  finds  himself  on  the  verge  of  voicing  a  feeling  which  is  obviously 
and  undeniably  true,  but  whicM  is  flatly  contradictory  to  the  conception 
of  himself  which  he  has  held.^fSee  62,  pp.  78-80,  for  a  striking  verbatim 
example  of  this  experience.]  Anxiety  results,  and  if  the  situation  is  ap- 
propriate (as  described  under  /)  this  anxiety  is  moderate,  and  the  result 
is  constructive.  But  if,  through  overzealous  and  effective  interpretation 
by  the  therapist,  or  through  some  other  means,  the  individual  is  brought 
face  to  face  with  more  of  his  denied  experiences  than  he  can  handle, 
disorganization  ensues  and  a  psychotic  break  occurs,  as  described  in 
statement  3.  We  have  known  this  to  happen  when  an  individual  has 
sought  c 'therapy"  from  several  different  sources  simultaneously.  It  has 
also  been  illustrated  by  some  of  the  early  experience  with  sodium 
pentathol  therapy.  Under  the  drug  the  individual  revealed  many  of  the 
experiences  which  hitherto  he  had  denied  to  himself,  and  which  ac- 
counted for  the  incomprehensible  elements  in  his  behavior.  Unwisely 
faced  with  the  material  in  his  normal  state  he  could  not  deny  its 
authenticity,  his  defensive  processes  could  not  deny  or  distort  the  ex- 
perience, and  hence  the  self-structure  was  broken,  and  a  psychotic  break 

230  CARL   R.    ROGERS 

Acute  psychotic  behaviors  appear  often  to  be  describable  as  behaviors 
which  are  consistent  with  the  denied  aspects  of  experience  rather  than 
consistent  with  the  self.  Thus  the  person  who  has  kept  sexual  impulses 
rigidly  under  control,  denying  them  as  an  aspect  of  self,  may  now  make 
open  sexual  overtures  to  those  with  whom  he  is  in  contact.  Many  of  the 
so-called  irrational  behaviors  of  psychosis  are  of  this  order. 

Once  the  acute  psychotic  behaviors  have  been  exhibited,  a  process 
of  defense  again  sets  in  to  protect  the  organism  against  the  exceedingly 
painful  awareness  of  incongruence.  Here  I  would  voice  my  opinion 
very  tentatively  as  to  this  process  of  defense.  In  some  instances  perhaps 
the  denied  experiences  are  now  regnant,  and  the  organism  defends  itself 
against  the  awareness  of  the  self.  In  other  instances  the  self  is  again 
regnant,  and  behavior  is  consistent  with  it,  but  the  self  has  been  greatly 
altered.  It  is  now  a  self  concept  which  includes  the  important  theme,  "I 
am  a  crazy,  inadequate,  unreliable  person  who  contains  impulses  and 
forces  beyond  my  control."  Thus  it  is  a  self  in  which  little  or  no  con- 
fidence is  felt. 

It  is  hoped  that  this  portion  of  the  theory  may  be  further  elaborated 
and  refined  and  made  more  testable  in  the  future. 

/.  The  Process  of  Reintegration 

In  the  situations  described  under  sections  G  and  H,  (and  probably  in 
situations  of  breakdown  as  described  under  I,  though  there  is  less  evidence 
on  this)  a  process  of  reintegration  is  possible,  a  process  which  moves  in  the 
direction  of  increasing  the  congruence  between  self  and  experience.  This 
may  be  described  as  follows : 

1.  In  order  for  the  process  of  defense  to  be  reversed — for  a  customarily 
threatening  experience  to  be  accurately  symbolized  in  awareness  and  as- 
similated into  the  self -structure.,  certain  conditions  must  exist. 

a.  There  must  be  a  decrease  in  the  conditions  of  worth. 

b.  There  must  be  an  increase  in  unconditional  self-regard. 

2.  The   communicated   unconditional  positive   regard  of  a  significant 
other  is  one  way  of  achieving  these"corrditions. 

a.  In  order  for  the  unconditional  positive  regard  to  be  communicated, 
it  must  exist  in  a  context  of  empathic  understanding. 

fe.  When  the  individual  perceives  such  unconditional  positive  regard, 
existing  conditions  of  worth  are  weakened  or  dissolved. 

c.  Another  consequence  is  the  increase  in  his  own  unconditional  positive 

d.  Conditions  2a  and  2b  above  thus  being  met,  threat  is  reduced,  the 
process  of  defense  is  reversed,  and  experiences  customarily  threatening 
are  accurately  symbolized  and  integrated  into  the  self  concept. 

Therapy,  Personality,  and  Interpersonal  Relationships  231 

3.  The  consequences  of  1  and  2  above  are  that  the  individual  is  less  likely 
to  encounter  threatening  experiences;  the  process  of  defense  is  less  frequent 
and  its  consequences  reduced;  self  and  experience  are  more  congruent;  self- 
regard  is  increased;  positive  regard  for  others  is  increased;  psychological  ad- 
justment is  increased;  the  organismic  valuing  process  becomes  increasingly 
the  basis  of  regulating  behavior;  the  individual  becomes  nearly  fully 

Comment.  This  section  is  simply  the  theory  of  therapy  which  we 
presented  earlier,  now  stated  in  a  slightly  more  general  form.  It  is  in- 
tended to  emphasize  the  fact  that  the  reintegration  or  restoration  of 
personality  occurs  always  and  only  (at  least  so  we  are  hypothesizing)  in 
the  presence  of  certain  definable  conditions.  These  are  essentially  the 
same  whether  we  are  speaking  of  formal  psychotherapy  continued  over 
a  considerable  period,  in  which  rather  drastic  personality  changes  may 
occur,  or  whether  we  are  speaking  of  the  minor  constructive  changes 
which  may  be  brought  about  by  contact  with  an  understanding  friend 
or  family  member. 

One  other  brief  comment  may  be  made  about  item  2a,  above.  Em- 
pathic  understanding  is  always  necessary  if  unconditional  positive  regard 
is  to  be  fully  communicated.  If  I  know  little  or  nothing  of  you,  and  ex- 
perience an  unconditional  positive  regard  for  you,  this  means  little  be- 
cause further  knowledge  of  you  may  reveal  aspects  which  I  cannot  so 
regard.  But  if  I  know  you  thoroughly,  knowing  and  empathically  under- 
standing a  wide  variety  of  your  feelings  and  behaviors,  and  still  ex- 
perience an  unconditional  positive  regard,  this  is  very  meaningful.  It 
comes  close  to  being  fully  known  and  fully  accepted. 

Specification  of  Functional  Relationships  in  the  Theory  of  Personality 

In  a  fully  developed  theory  it  would  be  possible  to  specify,  with 
mathematical  accuracy,  the  functional  relationships  between  the  several 
variables.  It  is  a  measure  of  the  immaturity  of  personality  theory  that  only 
the  most  general  description  can  be  given  of  these  functional  relation- 
ships. We  are  not  yet  in  a  position  to  write  any  equations.  Some  of  the 
relationships  implied  in  section  II  may  be  specified  as  follows: 

The  more  actualizing  the  experience,  the  more  adient  the  behavior 
(A5,  6). 

The  more  numerous  or  extensive  the  conditions  of  worth,  the  greater 
the  proportion  of  experience  which  is  potentially  threatening  (Fl,  2). 

The  more  numerous  or  extensive  the  conditions  of  worth,  the  greater 
the  degree  of  vulnerability  and  psychological  maladjustment  (F3). 

The  greater  the  proportion  of  experience  which  is  potentially  threat- 
ening, the  greater  the  probability  of  behaviors  which  maintain  and  en- 

232  CARL   R.    ROGERS 

hance  the  organism  without  being  recognized  as  self-experiences  (Gla, 

The  more  congruence  between  self  and  experience,  the  more  ac- 
curate will  be  the  symbolizations  in  awareness  (Gla,  and  HI,  2,  3). 

The  more  numerous  or  extensive  the  conditions  of  worth,  the  more 
marked  will  be  the  rigidity  and  inaccuracies  of  perception,  and  the 
greater  the  degree  of  intensionality  (#4) . 

The  greater  the  degree  of  incongruence  experienced  in  awareness, 
the  greater  the  likelihood  and  degree  of  disorganization  (73) . 

The  greater  the  degree  of  experienced  unconditional  positive  regard 
from  another,  based  upon  empathic  understanding,  the  more  marked 
will  be  the  dissolution  of  conditions  of  worth,  and  the  greater  the  pro- 
portion of  incongruence  which  will  be  eliminated  (/2,  3 ) . 

In  other  respects  the  relationships  in  section  /  have  already  been 
specified  in  the  theory  of  therapy. 

Evidence.  The  first  sections  of  this  theory  are  largely  made  up  of 
logical  constructs,  and  propositions  which  are  only  partly  open  to  em- 
pirical proof  or  disproof. 

Section  F  receives  some  confirmation  from  Cartwright  [9],  and  Diller 
[14],  Section  H  from  Chodorkoff  [10]  and  Cartwright  [9],  whereas 
Goldiamond  [22]  introduces  evidence  which  might  modify  the  definition 
of  subception.  Section  /  is  supported  by  the  evidence  previously  given 
for  the  theory  of  therapy  in  Part  I. 

Because  it  is  a  closely  reasoned  and  significant  experimental  testing 
of  certain  of  the  hypotheses  and  functional  relationships  specified  in  this 
portion  of  the  theory,  ChodorkofFs  study  [10]  will  be  described  briefly. 
His  definitions  were  taken  directly  from  the  theory.  Defensiveness,  for 
example,  is  defined  as  the  process  by  which  accurate  symbolizations  of 
threatening  experiences  are  prevented  from  reaching  awareness. 

He  concentrated  on  three  hypotheses  which  may  be  stated  in  theoreti- 
cal terms  as  follows : 

1.  The  greater  the  congruence  between  self  and  experience,  the  less 
will  be  the  degree  of  perceptual  defensiveness  exhibited. 

2.  The  greater  the  congruence  between  self  and  experience,  the  more 
adequate  will  be  the  personality  adjustment  of  the  individual,  as  this 
phrase  is  commonly  understood. 

3.  The  more  adequate  the  personality  adjustment  of  the  individual 
(as  commonly  understood),  the  less  will  be  the  degree  of  perceptual  de- 
fensiveness exhibited. 

Thus  it  will  be  seen  that  he  was  testing  one  of  the  definitions  of  the 
theory  (Congruence  equals  psychological  adjustment)  against  clinical 
and  common-sense  reality.  He  was  also  testing  one  of  the  relationships 
specified  by  the  theory  (Degree  of  congruence  is  inversely  related  to  de- 

Therapy,  Personality,  and  Interpersonal  Relationships  233 

gree  of  defensiveness) .  For  good  measure  he  also  completes  the  triangle 
by  testing  the  proposition  that  adjustment  as  commonly  understood  is  in- 
versely related  to  degree  of  defensiveness. 

He  gave  the  following  operational  meanings  to  the  essential  terms : 

1.  Self  is  defined  as  a  Q  sort  of  self-referent  items  sorted  by  the  in- 
dividual to  represent  himself  as  of  now. 

2.  Experience.  An  exact  matching  of  the  theoretical  meaning  with 
given  operations  is  of  course  difficult.  Chodorkoff  avoids  the  term  "ex- 
perience," but  operationally  defines  it  by  an  "objective  description" 
which  is  a  Q  sort  by  a  clinician  of  the  same  self-referent  items,  this  sort- 
ing being  based  on  a  thorough  clinical  knowledge  of  the  individual, 
gained  through  several  projective  tests.  Thus  the  total  experiencing  of 
the  individual,  as  distinct  from  the  self-concept  he  possesses  in  aware- 
ness, is  given  a  crude  operational  definition  by  this  means. 

3.  Perceptual  defensiveness  is  defined  as  the  difference  in  recogni- 
tion time  between  a  group  of  neutral  words  tachistoscopically  presented 
to  the  individual,  and  a  group  of  personally  threatening  words  similarly 
presented.  (The  selection  of  the  words  and  the  technique  of  presentation 
were  very  carefully  worked  out,  but  details  would  be  too  lengthy  here. ) 

4.  Personal   adjustment  as   commonly  understood  was  defined  as 
a  combined  rating  of  the  individual  by  four  competent  judges,  the  rating 
being  based  on  biographical  material,  projective  tests,  and  other  infor- 

These  definitions  provide  an  operational  basis  for  four  measures 
entirely  independent  of  one  another. 

Chodorkoff  translates  his  hypotheses  into  operational  predictions 
as  follows: 

1.  The  higher  the  correlation  between  the  individual's  self-sort  and 
the  clinician's  sorting  for  his  total  personality,  the  less  will  be  the  differ- 
ence in  his  recognition  threshold  between  neutral  and  threatening  words. 

2.  The  higher  the  correlation  between  the  self -sort  and  the  clinician's 
sorting  for  the  total  personality  the  higher  will  be  the  rating  of  personal 
adjustment  by  the  four  judges. 

3.  The  higher  the  adjustment  rating  by  the  four  judges,  the  lower 
will  be  the   difference  in  recognition  threshold  between  neutral  and 
threatening  words. 

All  three  of  these  predictions  were  empirically  upheld  at  levels  of 
statistical  significance,  thus  confirming  certain  portions  of  the  theory. 

This  study  illustrates  the  way  in  which  several  of  the  theoretical  con- 
structs have  been  given  a  partial  operational  definition.  It  also  shows 
how  propositions  taken  or  deduced  from  the  theory  may  be  empirically 
tested.  It  suggests,  too,  the  complex  and  remote  behavioral  predictions 
which  may  be  made  from  the  theory. 

234  CARL    R.    ROGERS 


Certain  directional  tendencies  in  the  individual  (ID  and  11-42)  and 
certain  needs  (IIC,  D)  have  been  explicitly  postulated  in  the  theory 
thus  far  presented.  Since  these  tendencies  operate  more  fully  under  cer- 
tain defined  conditions,  there  is  already  implicit  in  what  has  been  given 
a  concept  of  the  ultimate  in  the  actualization  of  the  human  organism. 
This  ultimate  hypothetical  person  would  be  synonymous  with  "the  goal 
of  social  evolution,"  "the  end  point  of  optimal  psychotherapy,"  etc.  We 
have  chosen  to  term  this  individual  the  fully  functioning  person. 

Although  it  contains  nothing  not  already  stated  earlier  under  I  and 
II,  it  seems  worthwhile  to  spell  out  this  theoretical  concept  in  its  own 

A.  The   individual   has  an  inherent   tendency  toward   actualizing  his 

B.  The  individual   has   the   capacity   and   tendency   to  symbolize   ex- 
periences accurately  in  awareness. 

1.  A  corollary  statement  is  that  he  has  the  capacity  and  tendency  to 
keep  his  self-concept  congruent  with  his  experience. 

C.  The  individual  has  a  need  for  positive  regard. 

D.  The  individual  has  a  need  for  positive  self-regard. 

E.  Tendencies  A  and  B  are  most  fully  realized  when  needs  C  and  D  are 
met.  More  specifically,  tendencies  A  and  B  tend  to  be  most  fully  realized 

1.  The  individual  experiences  unconditional  positive  regard  from  sig- 
nificant others. 

2.  The  pervasiveness  of  this  unconditional  positive  regard  is  made  evi- 
dent  through   relationships    marked    by    a    complete    and    communicated 
empathic  understanding  of  the  individual's  frame  of  reference. 

F.  If  the  conditions  under  E  are  met  to  a  maximum  degree,  the  in- 
dividual who  experiences  these  conditions  will  be  a  fully  functioning  person. 
The  fully  functioning  person  will  have  at  least  these  characteristics : 

1.  He  will  be  open  to  his  experience. 

a.  The  corollary  statement  is  that  he  will  exhibit  no  defensiveness. 

2.  Hence  all  experiences  will  be  available  to  awareness. 

3.  All  symbolizations  will  be  as  accurate  as  the  experiential  data  will 

4.  His  self-structure  will  be  congruent  with  his  experience. 

5.  His  self-structure  will  be  a  fluid  gestalt,  changing  flexibly  in  the 
process  of  assimilation  of  new  experience. 

6.  He  will  experience  himself  as  the  locus  of  evaluation. 

a.  The  valuing  process  will  be  a  continuing  organismic  one. 

Therapy,  Personality,  and  Interpersonal  Relationships  235 

7.  He  will  have  no  conditions  of  worth. 

a.  The  corollary  statement  is  that  he  will  experience  unconditional  self- 

8.  He  will  meet  each  situation  with  behavior  which  is  a  unique  and 
creative  adaptation  to  the  newness  of  that  moment. 

9.  He  will  find  his  organismic  valuing  a  trustworthy  guide  to  the  most 
satisfying  behaviors,  because 

a.  All  available  experiential  data  will  be  available  to  awareness  and  used. 

b.  No  datum  of  experience  will  be  distorted  in,  or  denied  to,  awareness. 

c.  The  outcomes  of  behavior  in  experience  will  be  available  to  awareness. 

d.  Hence  any  failure  to  achieve  the  maximum  possible  satisfaction,  be- 
cause of  lack  of  data,  will  be  corrected  by  this  effective  reality  testing. 

10.  He  will  live  with  others  in  the  maximum  possible  harmony,  because 
of  the  rewarding  character  of  reciprocal  positive  regard  (IIClc) . 

Comment.  It  should  be  evident  that  the  term  "the  fully  functioning 
person"  is  synonymous  with  optimal  psychological  adjustment,  optimal 
psychological  maturity,  complete  congruence,  complete  openness  to  ex- 
perience, complete  extensionality,  as  these  terms  have  been  defined. 

Since  some  of  these  terms  sound  somewhat  static,  as  though  such 
a  person  "had  arrived,"  it  should  be  pointed  out  that  all  the  character- 
istics of  such  a  person  are  process  characteristics.  The  fully  functioning 
person  would  be  a  person-in-process,  a  person  continually  changing. 
Thus  his  specific  behaviors  cannot  in  any  way  be  described  in  advance. 
The  only  statement  which  can  be  made  is  that  the  behaviors  would  be 
adequately  adaptive  to  each  new  situation,  and  that  the  person  would  be 
continually  in  a  process  of  further  self-actualization.  For  a  more  com- 
plete exposition  of  this  whole  line  of  thought  the  reader  may  wish  to  see 
my  paper  on  the  fully  functioning  person  [64] . 

Specification  of  Functions.  Our  present  state  of  thinking  can  be 
given  in  one  sentence.  The  more  complete  or  more  extensive  the  condi- 
tions Ely  E2,  the  more  closely  will  the  individual  approach  the  asymp- 
totic characteristics  Fl  through  FIQ. 

Evidence.  The  evidence  regarding  outcomes  of  therapy  is  in  a  gen- 
eral way  confirmatory  of  the  direction  taken  in  this  theory,  though  by 
its  very  nature  it  can  never  be  completely  tested,  since  it  attempts  to 
define  an  asymptote. 


The  most  recent  extension  of  our  theoretical  constructs  has  been  the 
attempt  to  formulate  the  order  which  appears  to  exist  in  all  interper- 
sonal relationships  and  interpersonal  communication.  This  formulation 

236  CARL    R.    ROGERS 

springs,  as  will  be  evident,  primarily  from  the  theory  of  therapy,  viewing 
the  therapeutic  relationship  as  simply  one  instance  of  interpersonal  rela- 
tionship. For  clarity  of  presentation  the  conditions,  process,  and  outcome 
of  a  deteriorating  relationship  and  a  deepening  or  improving  relationship 
will  be  set  forth  separately.  Actually  these  are  two  points  or  spaces  on  a 

A.  The  Conditions  of  a  Deteriorating  Relationship 

For  communication  to  be  reduced,  and  for  a  relationship  to  deteriorate, 
the  following  conditions  are  necessary: 

1.  A  person  Y  is  willing  to  be  in  contact  with  person  X  and  to  receive 
communication  from  him.    (Note:   Y's  characteristics  do  not  need  to  be 
specified,  beyond  saying  that  he  is  an  "average  person,53  with  some  malad- 
justment,  some   incongruence,   some    defensiveness.    The   theory   is   stated 
largely  in  terms  of  person  X. ) 

2.  Person  X  desires  (at  least  to  a  minimal  degree)   to  communicate  to 
and  be  in  contact  with  Y. 

3.  Marked  incongruence  exists  in  X  among  the  three  following  elements: 

a.  His  experience  of  the  subject  of  communication  with  Y.  (Which  may 
be  the  relationship  itself,  or  any  other  subject.) 

b.  The  symbolization  of  this  experience  in  his  awareness,  in  its  relation 
to  his  self-concept. 

c.  His   conscious  communicated  expression    (verbal  and/or  motor)    of 
this  experience. 

Comment.  If  the  discrepancy  in  3  is  a  vs.  b,  c,  then  X  is  psy- 
chologically maladjusted  in  this  respect,  and  the  immediate  consequences 
of  the  condition  tend  to  be  personal.  If  the  discrepancy  is  a,  b,  vs.  c, 
then  the  state  tends  to  be  labeled  deceit,  and  the  immediate  con- 
sequences tend  to  be  social. 

The  extreme  of  this  incongruence,  and  hence  one  end  point  of  the 
continuum,  would  be  a  complete  or  almost  complete  incongruence  or 
dissociation  between  the  experience,  its  cognitive  meaning  (symboliza- 
tion ) ,  and  its  expression. 

B.  The  Process  of  a  Deteriorating  Relationship 

When  the  preceding  conditions  exist  and  continue,  a  process  is  initiated 
which  tends  to  have  these  characteristics  and  directions : 

1.  The  communications  of  X  to  Y  is  contradictory  and/or  ambiguous, 

a.  Expressive  behaviors  which  are  consistent  with  X's  awareness  of  the 
experience  to  be  communicated. 

Therapy,  Personality,  and  Interpersonal  Relationships  237 

b.  Expressive  behaviors  which  are  consistent  with  those  aspects  of  the 
experience  not  accurately  symbolized  in  X5s  awareness.  (See  IIG 
above. ) 

2.  Y  experiences  these  contradictions  and  ambiguities. 

a.  He  tends  to  be  aware  only  of  Bla,  that  is  X's  conscious  communica- 

b.  Hence  his  experience  of  X's  communication  tends  to  be  incongruent 
with  his  awareness  of  same. 

c.  Hence  his  response  tends  also  to  be  contradictory  and/or  ambiguous, 
his  responses  having  the  same  qualities  described  for  X  in  Bla,  b. 

3.  Since  X  is  vulnerable,  he  tends  to  perceive  Y's  responses  as  potentially 

a.  Hence  he  tends  to  perceive  them  in  distorted  fashion,  in  ways  which 
are  congruent  with  his  own  self -structure. 

b.  Hence  he  is  inaccurate  in  his  perception  of  Y's  internal  frame  of 
reference,  and  does  not  experience  a  high  degree  of  empathy. 

c.  Because  Y  is  perceived  as  a  potential  threat,  X  cannot  and  does  not 
experience  unconditional  positive  regard  for  Y.   (Note:  thus  X  pro- 
vides the  reverse  of  the  conditions  for  therapy  as  described  in  1-43, 

4.  Y  experiences  himself  as  receiving  at  most  a  selective  positive  regard. 

5.  Y  experiences  a  lack  of  understanding  or  empathy. 

6.  The  more  Y  experiences  a  selectiveness  of  positive  regard  and  an 
absence  of  empathy,  the  less  free  he  is  to  express  feelings,  the  less  likely 
he  is  to  express  self -referent  feelings,  the  less  likely  he  is  to  be  extensional  in 
his  perceptions,  the  less  likely  he  is  to  express  incongruencies  between  self  and 
experience,  the  less  likely  he  is  to  reorganize  his  self -concept.  (Note:  in  gen- 
eral, the  process  of  personality  changes  as  described  in  IB  is  reversed.) 

7.  Since  Y  is  expressing  less  of  his  feelings,  X  is  even  more  unlikely  to 
perceive  Y's  internal  frame  of  reference  with  accuracy,  and  both  inaccuracy 
of  perception  and  distortion  of  perception  make  defensive  reactions  on  X's 
part  more  likely. 

8.  Another   characteristic   which   may  exist,   particularly  if  X's   com- 
munication is  primarily  of  negative  feelings,  is  that  those  aspects  of  ex- 
perience which  are  not  accurately  symbolized  by  X  in  his  awareness  tend, 
by  defensive  distortion  of  perception,  to  be  perceived  in  Y. 

9.  If  this  occurs,  Y  tends  to  be  threatened  to  the  degree  that  these  relate 
to  his  own  incongruences,  and  to  exhibit  defensive  behaviors. 

3  This  is  a  crucial  point.  If  Y  is  sufficiently  open  to  his  experience  that  he  is 
aware  of  X's  other  communication — described  in  Bib — then  b  and  c  below  do  not 
follow,  and  his  own  response  to  X  is  clear  and  congruent.  If  in  addition  to  his 
awareness  of  all  of  X's  communication  he  experiences  an  unconditional  positive 
regard  for  X,  then  this  would  become  an  improving  relationship,  as  described  in 
sections  D,  E,  and  F  which  follow. 

238  CARL    R.    ROGERS 

C.  The  Outcome  of  a  Deteriorating  Relationship 

The  continuance  of  this  process  results  in 

1.  Increased  defensiveness  on  the  part  of  X  and  Y. 

2.  Communication  which  is  increasingly  superficial,  expressive  of  less  of 
the  total  individual. 

3.  The  perceptions  of  self  and  others,  because  of  the  increased  defensive- 
ness,  are  organized  more  tightly. 

4.  Hence  incongruence  of  self  and  expression  remains  in  status  quo,  or 
is  increased. 

5.  Psychological  maladjustment  is  to  some  degree  facilitated  in  both. 

6.  The  relationship  is  experienced  as  poor. 

Comment  on  A,  B,  C.  It  may  clarify  this  technical  and  theoretical 
description  of  a  deteriorating  relationship  to  illustrate  it  from  some  com- 
monplace experience.  Let  us,  for  example,  take  the  relationship  of  a 
mother,  X,  toward  her  child,  Y.  There  is,  of  course,  mutual  willingness 
to  be  in  psychological  contact.  The  mother  feels  "You  annoy  me  because 
you  interfere  with  my  career,35  but  she  cannot  be  aware  of  this  because 
this  experience  is  incongruent  with  her  concept  of  herself  as  a  good 
mother.  Her  perception  of  this  experience  in  herself  is  distorted,  becoming 
"I  am  annoyed  at  this  instance  of  your  behavior.  I  love  you  but  I  must 
punish  you."  This  is  an  acceptable  symbolization  of  her  experience,  and 
it  is  this  which  she  consciously  communicates  to  the  child. 

But  Y  receives  not  only  this  conscious  communication.  He  also  ex- 
periences (but  tends  to  be  unaware  of)  the  expressive  behaviors  in- 
dicating a  more  general  dislike  of  himself.  His  response  may  be  of  several 
sorts,  but  its  essential  characteristic  is  that  it  will  express  the  incongruence 
which  her  divided  communication  has  set  up  in  him.  One  possibility  is 
that  he  will  experience  himself  as  bad  and  unloved,  even  when  his 
awareness  of  his  behavior  is  that  he  is  "good."  Hence  he  will  act  and 
feel  guilty  and  bad,  even  when  behaving  in  an  approved  manner.  This 
type  of  response  is  threatening  to  the  mother,  because  his  behaviors 
expressing  badness  and  unlovedness  threaten  to  bring  into  awareness 
her  own  rejecting  feelings.  Consequently  she  must  further  distort  her 
perception  of  his  behavior,  which  now  seems  to  her  "sneaky"  or  "hang- 
dog" as  well  as  being  occasionally  annoying.  The  more  this  cycle 
continues,  the  less  acceptance  Y  feels,  the  less  adequately  he  can  express 
his  feelings,  the  more  difficult  it  is  for  his  mother  to  achieve  any  empathic 
understanding,  the  more  completely  the  two  are  estranged  in  the  rela- 
tionship, the  more  maladjusted  each  becomes.  It  is  the  exact  steps  in 
such  a  relationship  which  we  have  endeavored  to  describe  in  the  three 
foregoing  sections — the  conditions  which  bring  it  about,  the  process 

Therapy^  Personality,  and  Interpersonal  Relationships  239 

by  which  deterioration  takes  place,  and  the  outcomes  of  such  a  de- 
teriorated relationship. 

D.  The  Conditions  of  an  Improving  Relationship 

For  communication  to  increase,  and  the  relationship  to  improve.,  the 
following  conditions  are  necessary: 

1.  A  person,  Y7,  is  willing  to  be  in  contact  with  person  X7,  and  to  receive 
communication  from  him. 

2.  Person  X'  desires  to  communicate  to  and  be  in  contact  with  Y7. 

3.  A  high  degree  of  congruence  exists  in  X'  between  the  three  following 
elements : 

a.  His  experience  of  the  subject  of  communication  with  Y/. 

b.  The  symbolization  of  this  experience  in  awareness  in  its  relation  to  his 
self -concept. 

c.  His  communicative  expression  of  this  experience. 

E.  The  Process  of  an  Improving  Relationship 

1.  The  communication  of  X'  to  Y'  is  characterized  by  congruence  of 
experience,  awareness.,  and  communication. 

2.  Y'  experiences  this  congruence  as  clear  communication.  Hence  his 
response  is  more  likely  to  express  a  congruence  of  his  own  experience  and 

3.  Since  X'  is  congruent  and  not  vulnerable  in  the  area  related  to  his 
communication,  he  is  able  to  perceive  the  response  of  Y'  in  an  accurate 
and  extensional  manner,  with  empathy  for  his  internal  frame  of  reference. 

4.  Feeling  understood,  Y7  experiences  some  satisfaction  of  his  need  for 
positive  regard. 

5.  X'  experiences  himself  as  having  made  a  positive  difference  in  the 
experiential  field  of  Y'. 

a.  Hence  reciprocally,  X7  tends  to  increase  in  feeling  of  positive  regard 
for  Y'. 

b.  Since  X'  is  not  vulnerable  in  the  area  of  the  communication,  the 
positive  regard  he  feels  for  Y7  tends  to  be  an  unconditional  positive 

6.  Y'  experiences  himself  in  a  relationship  which,  at  least  in  the  area  of 
communication,   is  characterized  by  congruence  on   the  part   of  X7,    an 
empathic  understanding  by  X7  of  the  internal  frame  of  reference,  and  an 
unconditional  regard.  (See  L43,  4,  5.) 

a.  Hence   all   the   characteristics   of  the  process  of  therapy    (IB}    are 
initiated,  within  the  confines  of  the  subject  of  communication. 

b.  Because  Y'  has  less  need  of  any  of  his  defenses  in  this  relationship,  any 
need  for  distortion  of  perception  is  decreased. 

c.  Hence  he  perceives  the  communications  of  X7  more  accurately. 

240  CARL   R.    ROGERS 

7.  Hence  communication  in  both  directions  becomes  increasingly  con- 
gruent, is  increasingly  accurately  perceived^  and  contains  more  reciprocal 
positive  regard. 

F.  Outcomes  of  an  Improving  Relationship 

The  continuance  of  this  process  results  in  the  following: 
1.  All  of  the  outcomes  of  therapy  (IC1  through  15)  may  occur,  subject 
to  the  time  limitation  of  the  relationship  between  X'  and  Y',  and  also  to 
the  mutually  understood  limitations  of  the  area  of  the  relationship  (e.g.,  it 
may  be  mutually  understood  that  it  is  only  a  lawyer-client  relationship,  or 
only  a  teacher-pupil  relationship,  thus  tending  to  exclude  many  areas  of 
expression  and  hence  to  that  degree  limiting  the  extent  of  the  outcomes) . 
Thus,  within  these  limitations,  the  relationship  facilitates  improved  con- 
gruence and  psychological  adjustment  in  both  X'  and  Yy. 

G.  A  Tentative  Law  of  Interpersonal  Relationships 

Taking  all  of  this  section,  we  may  attempt  to  compress  it  into  one  over- 
all law  governing  interpersonal  relationships,  specifying  the  functional  rela- 
tionship between  the  constructs.  Here  is  such  an  attempt. 

Assuming  a  minimal  mutual  willingness  to  be  in  contact  and  to  receive 
communications,  we  may  say  that  the  greater  the  communicated  congruence 
of  experience^  awareness,  and  behavior  on  the  part  of  one  individual,  the 
more  the  ensuing  relationship  will  involve  a  tendency  toward  reciprocal 
communication  with  the  same  qualities,  mutually  accurate  understanding 
of  the  communications,  improved  psychological  adjustment  and  functioning 
in  both  parties,  and  mutual  satisfaction  in  the  relationship. 

Conversely,  the  greater  the  communicated  incongruence  of  experience, 
awareness,  and  behavior,  the  more  the  ensuing  relationship  will  involve 
further  communication  with  the  same  quality,  disintegration  of  accurate 
understanding,  lessened  psychological  adjustment  in  both  parties,  and 
mutual  dissatisfaction  in  the  relationship. 

Comment.  This  is  still  a  theory  in  the  making,  rather  than  a  finished 
product.  It  does  not  grow  out  of  consideration  of  research  data  and 
grows  only  partly  out  of  experience.  Basically,  it  is  deduced  from  the 
theory  of  therapy  and  projects  into  a  new  area  a  series  of  hypotheses 
which  now  require  confirmation  or  disproof.  The  evidence  gained  in  such 
studies  should  not  only  modify  or  confirm  the  theory  of  interpersonal 
relationships  but  should  reflexively  throw  new  light  on  the  theory  of  ther- 
apy as  well. 

Evidence.  It  is  believed  that  there  is  evidence  from  experience  and 
some  research  evidence  concerning  this  theory.  It  seems  preferable,  how- 
ever, simply  to  present  it  as  a  deduced  theory. 

Therapy,  Personality,  and  Interpersonal  Relationships  241 


To  spell  out  in  detail  the  various  theories  of  application  which  have 
been  partially  developed,  would  be  too  repetitious  of  what  has  gone 
before.  Hence  only  a  descriptive  suggestion  will  be  given  in  each  area 
of  the  aspects  of  theory  which  would  be  applicable. 

Family  life.  The  theoretical  implications  would  include  these: 

1.  The  greater  the  degree  of  unconditional  positive  regard  which  the 
parent  experiences  toward  the  child : 

a.  The  fewer  the  conditions  of  worth  in  the  child. 

b.  The  more  the  child  will  be  able  to  live  in  terms  of  a  continuing 
organismic  valuing  process. 

c.  The  higher  the  level  of  psychological  adjustment  of  the  child. 

2.  The  parent  experiences  such  unconditional  positive  regard  only  to  the 
extent  that  he  experiences  unconditional  self-regard. 

3.  To   the   extent   that   he   experiences   unconditional  self-regard,   the 
parent  will  be  congruent  in  the  relationship. 

a.  This  implies  genuineness  or  congruence  in  the  expression  of  his  own 
feelings  (positive  or  negative). 

4.  To   the   extent  that  conditions   1,  25  and  3   exist,   the  parent  will 
realistically   and    empathically   understand   the    child's    internal  frame   of 
reference  and  experience  an  unconditional  positive  regard  for  him. 

5.  To  the  extent  that  conditions  1  through  4  exist,  the  theory  of  the 
process  and  outcomes  of  therapy  (IB,  C) ,  and  the  theory  of  the  process  and 
outcomes  of  an  improving  relationship  (I'VE,  F) ,  apply. 

Comment.  Stated  thus  briefly,  the  applications  to  family  life  may 
easily  be  misunderstood.  For  a  presentation  of  these  and  related  ideas, 
the  reader  is  referred  to  [65]. 

Education  and  learning.  To  the  extent  that  education  is  concerned 
with  learnings  which  significantly  influence  behavior  and  facilitate 
change  in  personality,  then  the  conditions  of  therapy  (1-4)  and  the 
conditions  of  an  improving  relationship  (IVD)  apply.  This  leads,  among 
other  things,  to  more  realistic,  accurate,  and  differentiated  perceptions 
(IC1,  2)  and  to  more  responsible  basing  of  behavior  upon  these  per- 
ceptions (IC3,  10,  15). 

Comment.  Since  a  reasonably  full  statement  of  the  theory  of 
facilitating  learning  has  already  been  set  forth  [62,  chap.  9],  no  at- 
tempt will  be  made  to  spell  it  out  in  detail  here,  even  though  a  number 
of  the  terms  and  constructs  in  this  earlier  presentation  are  not  precisely 
those  which  are  used  here. 

242  CARL    R.    ROGERS 

Evidence.  Several  studies  of  the  application  of  this  theory  to  the 
educational  process  have  been  made.  Gross  [26],  Schwebel  and  Asch 
[74].,  Asch  [3],  and  Faw  [17,  18],  supply  evidence  which  in  general  is 

Group  leadership.  Building  upon  the  postulate  regarding  the  nature 
of  the  individual  (ID)  and  extending  this  to  apply  to  groups,  it  has  been 
hypothesized  that  to  the  extent  that  a  perceived  leader  provides  the 
conditions  of  therapy  ( L43,  43  5 )  or  of  an  improving  relationship  ( TVD } , 
certain  phenomena  will  occur  in  the  group.  Among  these  are  the  follow- 
ing: the  perceptual  resources  of  the  group  will  be  more  widely  used, 
more  differentiated  data  will  be  provided  by  the  group,  thinking  and 
perceptions  will  become  more  extensional,  self-responsible  thinking  and 
action  will  increase,  a  greater  degree  of  distributive  leadership  will  de- 
velop, and  there  will  be  more  effective  long-range  problem  solving.  All 
of  these  consequences  flow  logically  from  the  theory  thus  far  presented. 
In  two  major  expositions  [24,  23],  Gordon  has  set  forth  carefully  the 
theory  of  application  in  this  field,  and  it  will  not  be  repeated  here.  The 
reader  is  referred  to  these  presentations  for  more  detail. 

Evidence.  The  studies  by  Roethlisberger  and  Dickson  [57],  Coch 
and  French  [11],  Radke  and  Klisurich  [53],  Gordon,  and  others  supply 
some  confirmatory  evidence  of  different  aspects  of  the  theory. 

Group  tension  and  conflict.  In  serious  situations  of  group  conflict, 
the  conditions  of  a  deteriorating  interpersonal  relationship  (IV A) 
usually  exist.  Drawing  both  from  the  theory  of  therapy  and  the  theory 
of  interpersonal  relationships,  certain  hypotheses  have  been  formulated 
in  regard  to  such  situations.  Since  these  introduce  a  somewhat  new 
point,  they  will  be  formulated  in  more  detail. 

For  our  present  purpose  we  may  assume  as  given  a  group  situation 
in  which  the  conditions  of  a  deteriorating  relationship  (IV A)  already 
exist,  with  defensive  behaviors  and  expressions  being  mutually  increased 
between  X  and  Y  and  Z,  different  members  of  the  group,  or  between 
different  subgroups  represented  by  X,  Y,  and  Z. 

A.  Conditions  of  Reduction  in  Group  Conflict 

Group  conflict  and  tension  will  be  reduced  if  these  conditions  exist. 

1.  A  person   (whom  we  term  a  facilitator)    is  in  contact  with  X,  Y, 

2.  The  facilitator  is  congruent  within  himself  in  his  separate  contacts 
with  X,  Y,  and  Z. 

3.  The  facilitator  experiences  toward  X,  Y,  and  Z,  separately: 

a.  An  unconditional  positive  regard.,  at  least  in  the  area  in  which  the 
members  of  the  group  are  communicating. 

Therapy,  Personality,  and  Interpersonal  Relationships  243 

b.  An  em  pat  hie  understanding  of  the  internal  frame  of  reference  of 
X,  Y,  Z,  at  least  In  the  area  in  which  the  members  of  the  group  are 

4.  X,  Y,  and  Z  perceive,  at  least  to  a  minimal  degree^  conditions  3a  and 
36.  (This  is  generally  because  3b  is  communicated  verbally.) 

B.  The  Process  of  Reduction  of  Group  Conflict 

If  the  above  conditions  exist  and  continue,  then : 

1.  The  various  elements  of  the  process  of  therapy  (IB)   take  place  to 
some  degree,  at  least  within  the  area  involved  in  the  group  communication. 

a.  One  of  the  important  elements  of  this  process  is  the  increase  in  dif- 
ferentiated perceptions  and  in  extensionality. 

b.  Another  important  element  is  the  reduction  of  threat  (see  IBS,  8a) 
in  the  experience  of  X,  Y,  Z. 

2.  Consequently  the  communications  of  Y  to  X  or  Z  to  X.,  are  less  de- 
fensive, and  more  nearly  congruent  with  the  experience  of  Y,  and  with  the 
experience  of  Z. 

3.  These  communications  are  perceived  with  increasing  accuracy  and 
extensionality  by  X. 

a.  Consequently  X  experiences  more  empathic  understanding  of  Y 
and  Z. 

4.  Because   he   is   experiencing   less   threat  from  Y  and   Z   and  more 
empathy  with  their  internal  frame  of  reference: 

a.  X  now  symbolizes  in  awareness  incongruencies  which  formerly  existed 
between  experience  and  awareness. 

b.  Consequently   his    defensive    distortions   of   his   own   experience   are 

c.  Hence  his  communication  to  Y  and  Z  becomes  a  more  extensional  ex- 
pression of  his  own  total  experience  in  regard  to  the  area  of  com- 

5.  The  conditions  now  exist  for  the  process  of  an  improving  relationship, 
and  the  phenomena  described  in  TVE  occur. 

Comment.  A  more  general  statement  of  the  views  presented  here 
theoretically  will  be  found  in  two  previous  papers  [63,  61].  This  theory 
is  a  deduction  from  the  theory  of  therapy,  and  the  theory  of  interpersonal 

Evidence.  Although  clinical  evidence  tends  to  confirm  the  theory  in 
small  face-to-face  groups,  and  Axline  [5]  has  given  an  account  of  such  a 
clinical  situation,  there  is  as  yet,  I  believe,  no  research  evidence  bearing 
on  this  aspect  of  the  theory.  Particularly  crucial  and  important  from  a 
social  point  of  view  will  be  investigations  involving  different  sizes  of 
groups.  Even  if  the  theory  is  fully  confirmed  in  small  f ace-to-face  groups, 
will  it  hold  true  in  larger  groups  where  communication  is  not  face-to- 

244  CARL   R.   ROGERS 

face?  There  is  also  a  question  involving  groups  composed  of  spokesmen, 
or  representatives,  where  the  individual  feels  that  he  cannot  speak  out  of 
his  own  experience  and  feeling,  but  only  in  a  way  dictated  by  his  con- 
stituents, who  are  not  present.  It  is  quite  clear  that  the  theory,  as 
formulated  here,  would  not  directly  apply  to  this  last  type  of  situation. 


Our  presentation  of  the  theoretical  system  is  completed.  It  is  to  be 
hoped  that  the  presentation  has  made  it  clear  that  this  is  a  developing 
system,  in  which  some  of  the  older  portions  are  being  formulated  with 
considerable  logical  rigor,  while  newer  portions  are  more  informal,  and 
contain  some  logical  and  systematic  gaps  and  flaws,  and  still  others  (not 
presented)  exist  as  highly  personal  and  subjective  hunches  in  the  minds 
of  members  of  the  client-centered  group.  It  is  also  to  be  hoped  that  it  is 
evident  that  this  is  a  system  which  is  in  a  continual  state  of  modification 
and  clarification.  Comparison  of  the  theory  as  given  above  with  the 
theory  of  therapy  and  personality  given  in  Client-centered  Therapy  in 
1951  [62,  chaps.  4,  11]  or  with  the  paper  presented  to  the  APA  in  1947 
[60]  will  show  that  although  the  major  directions  have  not  markedly 
changed,  there  have  been  many  changes  in  the  constructs  employed,  and 
far-reaching  changes  in  the  organization  of  the  theory.  This  ongoing 
process  of  revision  is  expected  to  continue. 

The  major  usefulness  of  the  systematic  theoretical  thinking,  aside 
from  the  personal  satisfaction  it  has  given,  has  been  the  stimulation  of 
research.  In  this  respect  there  seems  little  doubt  that  it  has  had  con- 
siderable success.  By  and  large  the  order  of  events  seems  to  have  been 
this — clinical  therapeutic  experience,  formulation  of  theory,  research 
which  tests  the  theory,  new  aspects  of  experience  perceived  because  of 
the  research,  modification  of  the  theory  in  the  light  of  the  new  experience 
and  the  research,  further  empirical  testing  of  the  revised  hypotheses. 

It  would  take  too  much  space  to  review  or  even  list  the  studies  which 
have  been  made.  This  would  also  be  an  unnecessary  duplication  since 
Seeman  and  Raskin  [77]  have  written  a  thoughtful  analysis  and  criticism 
of  55  of  the  research  studies  in  therapy  and  personality  which  have  been 
stimulated  by  this  point  of  view  and  completed  during  the  years  1942- 
195 1.4  Suffice  it  to  say  that  clusters  of  research  investigations  have  been 
made  around  each  of  the  following  subjects  of  inquiry : 

1.  The  events  and  process  of  therapy.  Analysis  of  recorded  thera- 
peutic interviews  in  terms  of  theoretical  constructs  has  been  a  major 
tool  here. 

4 Since  writing  the  above  D.  S.  Cartwright  has  published:  Annotated  bibliog- 
raphy of  research  and  theory  construction  in  client-centered  therapy,  /.  counsel. 
PsychoL,  1957,  4,  82-100. 

Therapy,  Personality ,  and  Interpersonal  Relationships  245 

2.  The  results  or  outcomes  of  therapy.   Tests  of  personality  and 
measures  of   different  aspects  of  behavior  have  been  the   major  in- 

3.  Investigation  of  personality  theory.   Hypotheses  regarding  per- 
ception of  self,  others,  external  reality,  and  perceived  locus  of  evaluation 
have  been  investigated  with  a  wide  range  of  instruments, 

4.  Application  of  theory  in  specific  fields.  Investigations  particularly 
in  the  facilitation  of  learning  and  in  group  leadership. 

Since  1951,  many  more  studies  have  been  completed  in  the  out- 
comes of  therapy,  an  important  collection  of  these  being  gathered  in 
Psychotherapy  and  Personality  Change  [70].  In  these  studies  the  prob- 
lem of  a  control  group  is  much  more  adequately  handled  than  heretofore, 
giving  the  findings  a  solidity  which  is  noteworthy.  If  the  reader  wishes 
to  obtain  a  first-hand  grasp  of  the  way  in  which  refinements  of  instru- 
mentation and  general  scientific  sophistication  have  developed  in  this 
field,  he  should  compare  the  seven  studies  of  therapeutic  outcome 
published  in  the  Journal  of  Consulting  Psychology  in  1949  (the  entire 
July  issue,  pp.  149-220)  with  the  thirteen  studies  published  in  Psy- 
chotherapy and  Personality  Change  ( 1954) . 

In  addition  to  the  many  studies  of  outcome  there  are  an  increasing 
number  which  have  as  their  primary  purpose  the  investigation  of  em- 
pirical predictions  made  from  personality  theory.  The  study  of  Ghodor- 
koff  [10],  already  cited,  is  an  excellent  example  of  this  group.  There  are 
also  studies  now  in  progress  which  draw  their  hypotheses  from  an  inte- 
gration of  the  theory  of  therapy  with  a  theory  of  perception  or  a  theory 
of  learning.  Such  studies  will,  it  is  hoped,  link  the  findings  in  the  field 
of  therapy  to  the  findings  in  older  and  more  established  fields  of 

The  bases  of  stimulation  of  research.  There  are,  in  the  writer's 
opinion,  several  basic  reasons  why  this  theoretical  system  has  been  help- 
ful in  giving  impetus  to  a  wide  variety  of  research  investigations. 

The  first  is  the  orienting  attitude  mentioned  in  the  first  section  of 
this  document,  that  scientific  study  can  begin  anywhere,  at  any  level 
of  crudity  or  refinement,  that  it  is  a  direction,  not  a  fixed  degree  of  in- 
strumentation. From  this  point  of  view,  a  recorded  interview  is  a  small 
beginning  in  scientific  endeavor,  because  it  involves  greater  objectification 
than  the  memory  of  an  interview;  a  crude  conceptualization  of  therapy 
and  crude  instruments  for  measuring  these  concepts,  are  more  scientific 
than  no  such  attempts.  Thus  individual  research  workers  have  felt  that 
they  could  begin  to  move  in  a  scientific  direction  in  the  areas  of  greatest 
interest  to  them.  Out  of  this  attitude  has  come  a  series  of  instruments  of 
increasing  refinement  for  analyzing  interview  protocols,  and  significant 
>  beginnings  have  been  made  in  measuring  such  seemingly  intangible  con- 

246  CARL    R.    ROGERS 

structs  as  the  self-concept  and  the  psychological  climate  of  a  therapeutic 

This  leads  me  to  what  I  believe  to  be  the  second  major  reason  for 
the  degree  of  success  the  theory  has  had  in  encouraging  research.  The 
constructs  of  the  theory  have,  for  the  most  part,  been  kept  to  those  which 
can  be  given  operational  definition.  This  has  seemed  to  meet  a  very 
pressing  need  for  psychologists  and  others  who  have  wished  to  advance 
knowledge  in  the  field  of  personality  but  who  have  been  handicapped  by 
theoretical  constructs  which  cannot  be  defined  operationally.  Take,  for 
example,  the  general  phenomena  encompassed  in  such  terms  as  the  self, 
the  ego,  the  person.  If  a  construct  is  developed — as  has  been  done — 
which  includes  those  inner  events  not  in  the  awareness  of  the  individual 
as  well  as  those  in  awareness,  then  there  is  no  satisfactory  way  at  the 
present  time  to  give  such  a  construct  an  operational  definition.  But  by 
limiting  the  self-concept  to  events  in  awareness,  the  construct  can  be 
given  increasingly  refined  operational  definition  through  the  Q  technique, 
the  analysis  of  interview  protocols,  etc.,  and  thus  a  whole  area  of  in- 
vestigation is  thrown  open.  In  time  the  resulting  studies  may  make  it 
possible  to  give  operational  definition  to  the  cluster  of  events  not  in 

The  use  of  operationally  definable  constructs  has  had  one  other 
effect.  It  has  made  completely  unnecessary  the  use  of  "success"  and 
"failure33— two  terms  which  have  no  scientific  usefulness — as  criteria  in 
studies  of  therapy.  Predictions  can  instead  be  made  in  terms  of  opera- 
tionally definable  constructs,  and  these  predictions  can  be  confirmed  or 
disconfirmed,  quite  separately  from  any  value  judgments  as  to  whether 
the  change  represents  "success"  or  "failure."  Thus  one  of  the  major 
barriers  to  scientific  advance  in  this  area  has  been  removed. 

A  third  and  final  reason  for  whatever  effectiveness  the  system  has 
had  in  mediating  research  is  that  the  constructs  have  generality.  Because 
psychotherapy  is  such  a  microcosm  of  significant  interpersonal  relation- 
ship, significant  learning,  and  significant  change  in  perception  and  in 
personality,  the  constructs  developed  to  order  the  field  have  a  high 
degree  of  pervasiveness.  Such  constructs  as  the  self-concept,  or  the  need 
for  positive  regard,  or  the  conditions  of  personality  change,  all  have 
application  to  a  wide  variety  of  human  activities.  Hence  such  constructs 
may  be  used  to  study  areas  as  widely  variant  as  industrial  or  military 
leadership,  personality  change  in  psychotic  individuals,  the  psychological 
climate  of  a  family  or  a  classroom,  or  the  interrelation  of  psychological 
and  physiological  change. 

The  problem  of  measurement  and  quantification.  I  do  not  feel  com- 
petent to  discuss,  at  a  sophisticated  level  of  statistical  knowledge,  the 
problems  of  measurement  which  have  been  met  by  our  group.  This  is 

Therapy,  Personality,  and  Interpersonal  Relationships  247 

best  left  to  others.  I  will  only  mention  three  examples  of  the  continuing 
trend  toward  ever  more  refined  quantification  of  the  data  of  psycho- 
therapy and  personality. 

The  researches  which  have  taken  their  start  from  client-centered 
theory  have  significantly  advanced  the  field  of  analysis  of  verbal 
protocols.  Working  with  recorded  interviews,  increasingly  exact  methods 
have  been  devised,  so  that  reliability  of  categorization  is  high,  and  very 
subtle  constructs,  such  as,  for  example,  an  "emergent  self-perception" 
can  be  objectified  and  measured.  The  attempt  has  been  made  by 
Grummon  [27]  to  integrate  some  of  the  methods  we  have  developed  with 
the  more  formal  methods  of  language  analysis. 

Other  research  workers  have  taken  the  Q  technique  as  developed 
by  Stephenson  [81],  and  have  exploited  it  in  a  variety  of  ways.  It  has 
been  used  to  give  an  operational  definition  to  the  self-concept,  to  pro- 
vide objectifications  of  a  diagnostician's  perception  of  an  individual  im- 
mediately comparable  to  that  individual's  self-perception,  to  measure 
the  quality  of  a  relationship  as  perceived  by  the  two  participants,  and  to 
test  a  variety  of  hypotheses  growing  from  personality  theory. 

Butler  [6]  has  developed  a  new  method  for  discovering  the  order 
which  exists  in  such  material  as  interview  protocols.  A  number  of  people 
working  with  him  have  begun  to  apply  this  method — termed  Rank 
Pattern  Analysis — to  problems  of  complex  analysis  which  hitherto  had 
been  baffling. 

Thus  in  a  number  of  different  areas  the  researches  stimulated  by 
client-centered  theory  have  not  only  contributed  to  the  empirical  base 
of  the  theory,  but  have  contributed  to  the  development  of  methodology 
as  well.  In  principle  there  seems  no  limit  to  the  refinement  of  measure- 
ment in  the  areas  covered  by  the  theory.  The  major  obstacle  to  progress 
has  been  the  lack  of  sufficient  inventiveness  to  develop  tools  of  measure- 
ment adequate  for  the  tasks  set  by  the  theory. 

Incompatible  evidence.  Some  of  the  evidence  related  to  the  theory 
has  been  cited  in  each  section.  It  will  have  been  noted  that  nearly  all 
of  this  evidence  has  been  confirmatory  and  that  which  is  not  confirming 
has  tended  to  be  confused.  There  is  almost  no  research  evidence  which 
appears  flatly  to  contradict  the  predictions  from  the  theory. 

Two  related  exceptions  are  the  study  reported  by  Carr  [8],  and 
a  portion  of  the  study  made  by  Grummon  and  John  [28,  also  37]  which 
is  discussed  by  Vargas  [85].  Briefly,  the  facts  seem  to  be  that  Carr 
and  John  had  pre-  and  posttherapy  projective  tests  analyzed  by  psy- 
chologists who  were  basically  diagnosticians.  They  found  little  or  no 
change  in  the  degree  of  adjustment,  in  the  projective  material.  In  a 
series  of  10  cases,  the  John  ratings  as  discussed  by  Vargas  had  a 
significant  negative  correlation  with  counselor  ratings.  Yet  when  these 

248  CARL   R.   ROGERS 

same  materials  are  analyzed  "blind33  by  therapeutically  oriented  re- 
searchers (for  example,  Dymond)  positive  change  is  found,  and  the 
correlation  with  counselor  ratings  is  significantly  positive. 

The  explanation  suggested  by  Vargas  is  that  the  diagnostician  tends 
to  think  of  adjustment  as  stability,  a  more  or  less  fixed  "level  of  de- 
fense" which  is  socially  acceptable.  The  therapeutically  oriented  worker 
— especially  if  influenced  by  client-centered  theory — tends  to  think  of 
psychological  adjustment  as  an  openness  to  experience,  a  more  fluid 
expressiveness  and  adaptiveness.  Hence  what  the  diagnostician  perceives 
as  loss  of  control  or  even  disorganization  may  be  perceived  by  the 
therapeutically  oriented  person  as  progress  toward  reduced  defensiveness 
and  greater  openness  to  experience.  How  deep  this  contradiction  goes, 
and  its  full  implications,  can  only  be  evaluated  in  the  light  of  further 

The  main  source  of  incompatible  evidence  is  not  research  evidence, 
but  a  clinical  point  of  view.  By  and  large  the  psychoanalytically  oriented 
Freudian  group  has  developed,  out  of  its  rich  clinical  experience,  a  point 
of  view  which  is  almost  diametrically  opposed  to  the  hypotheses  regard- 
ing the  capacities  and  tendencies  of  the  human  organism  formulated 
above  in  Dl,  2,  3,  and  also  diametrically  opposed  to  the  theory  of  the 
fully  functioning  person  in  III.  Very  briefly  stated,  the  Freudian  group, 
on  the  basis  of  its  experience,  tends  to  see  the  individual  as  "innately 
destructive53  (to  use  Karl  Menninger's  words)  and  hence  in  need  of 
control.  To  members  of  this  group  the  hypothetical  individual  pictured 
earlier  under  A  Theory  of  the  Fully  Functioning  Person  is  a  psychopathic 
personality,  because  they  see  nothing  that  would  control  him.  The 
hypothesis  that  self-control  would  be  natural  to  the  person  who  is  with- 
out defenses  appears  to  them  untenable. 

In  very  much  related  fashion,  the  theory  which  Gordon  and  others 
have  formulated  regarding  group  behavior  and  group  leadership  is 
almost  diametrically  opposed  to  the  Freudian  theory  in  this  respect. 
Freud's  statements  that  "groups  have  never  thirsted  after  truth53  and 
that  "a  group  is  an  obedient  herd  which  could  never  live  without  a 
master3'  suggests  something  of  the  deep  discrepancy  which  exists  between 
the  two  views. 

Though  the  psychoanalytic  theory  in  these  two  respects  is  not  sup- 
ported by  any  research  evidence,  it  nevertheless  deserves  serious  con- 
sideration because  of  the  soil  of  clinical  experience  out  of  which  it 
originally  grew.  The  discrepancy  seems  even  more  puzzling  and  challeng- 
ing when  it  is  realized  that  both  the  Freudian  group  and  the  client- 
centered  group  have  developed  their  theories  out  of  the  deep  and 
intimate  personal  relationships  of  psychotherapy. 

It  is  my  belief  that  the  discrepancy  can  be  understood  in  a  way 

Therapy,  Personality,  and  Interpersonal  Relationships  249 

which  leaves  the  client-centered  theory  intact,  but  this  does  not  seem 
to  be  the  place  for  such  a  discussion.  It  seems  best  to  present  these 
incompatible  views  for  what  they  are,  two  theoretical  stands  which  are 
in  flat  contradiction  on  some  basic  points.  Only  new  integrations  of 
theory  and  much  deeper  research  investigations  can  resolve  the  difference. 

A  continuing  program  of  theory  and  research.  The  theoretical 
system  and  the  research  program  which  are  connected  with  client- 
centered  therapy  have  grown  from  within  themselves.  This  point  can 
hardly  be  overemphasized.  The  thought  that  we  were  making  a  start 
on  a  theoretical  system  would  for  me  have  been  a  most  distasteful  notion 
even  as  little  as  a  dozen  years  ago.  I  was  a  practical  clinician  and  held 
(horrible  dictu!}  an  open  scorn  of  all  psychological  theory,  as  my  early 
students  at  Ohio  State  can  testify.  This  was  true  even  at  the  same  time 
that  I  was  beginning  to  discern  the  orderliness  which  existed  in  the 
therapeutic  process.  I  like  to  think  that  the  theoretical  system  and  far- 
reaching  web  of  research  which  have  developed,  have  grown  in  an 
organic  fashion.  Each  plodding  step  has  simply  been  a  desire  to  find  out 
this,  a  desire  to  find  out  that,  a  need  for  perceiving  whatever  con- 
sistencies, or  invariances,  or  order  exists  in  the  material  thus  far  un- 

Consequently  when  I  am  asked,  as  I  am  in  the  outline  suggested  for 
this  paper,  "the  extent  to  which  the  systematic  program  has  been 
realized,"  I  feel  it  is  the  wrong  question  for  this  system.  I  have  no  idea 
what  will  be  the  ultimate  realization  of  the  living  program  which  has 
developed.  I  can  see  some  of  the  likely  next  steps,  or  the  current  di- 
rections, but  have  no  assurance  that  these  will  be  taken.  We  have  con- 
tinued to  move  in  the  directions  which  are  experienced  as  rewarding,  not 
necessarily  in  those  directions  which  logic  points  out.  I  believe  this  has 
been  the  strength  of  the  program,  and  I  trust  it  will  continue. 

Thus  I  believe  that  we  are  likely  to  see  progress  in  the  following 
directions,  but  I  am  not  sure  of  any  of  them.  It  seems  likely  that  further 
moves  will  be  made  toward  theory  and  research  in  the  field  of  perception, 
enriching  that  field  by  the  insights  gained  in  therapy,  and  being  enriched 
by  the  wealth  of  research  data  and  theory  in  perception  which  can  be 
brought  to  bear  in  the  refinement  of  the  theories  we  are  developing. 
One  such  study  now  in  progress,  for  example,  is  attempting  to  investigate 
perceptual  changes  which  occur  during  therapy.  The  measures  range 
from  those  entirely  concerned  with  social  perception — of  people,  of  rela- 
tionships— to  those  entirely  concerned  with  the  physical  perception  of 
form,  color,  and  line.  Does  therapy  change  only  social  perception,  or 
does  it  alter  even  the  most  basic  perceptual  processes?  If  not,  where  on 
this  continuum  does  change  cease  to  occur? 

I  visualize  the  same  type  of  rapprochement  with  learning  theory, 

250  CARL   R.    ROGERS 

where  in  my  judgment  we  have  much  to  offer  in  the  way  of  new  di- 
rections in  that  field,  as  well  as  being  able  to  use  much  of  the  material 
available  there.  It  also  seems  likely  that  a  number  of  the  hypotheses  we 
are  formulating  may  be  tested  in  the  laboratory,  some  on  human  and 
some  on  animal  subjects,  thus  linking  the  field  of  personality  and  therapy 
with  so-called  experimental  psychology.  There  seems  no  reason,  for  ex- 
ample, why  research  on  the  establishment  and  consequences  of  condi- 
tions of  worth,  as  spelled  out  in  this  theory,  might  not  be  carried  out  on 
higher  animals,  with  a  wider  range  of  experimental  conditions  and  more 
adequate  controls  than  could  be  achieved  with  human  subjects. 

I  regard  it  as  possible  that  there  may  be  a  closer  linking  of  our 
theory  with  the  developing  interest  in  creativity  in  the  humanities  and 
social  sciences  generally,  and  I  trust  that  this  theory  may  provide  a 
number  of  relevant  hypotheses  for  testing.  I  regard  it  as  veiy  likely  that 
the  implications  of  this  body  of  theory  for  industrial  production  will  be 
developed  much  more  fully — the  beginnings,  as  described  by  Richard 
in  Gordon's  book  [23],  seem  very  exciting.  I  believe  it  is  possible  that 
the  near  future  may  see  a  clear  linking  with  the  psychiatric  group  and  a 
testing  of  the  theory  in  a  wider  variety  of  human  disorders,  with  a  re- 
duction in  the  professional  parochialism  which  has  thus  far  kept  the 
medical  group  largely  ignorant  of  the  research  in  this  field. 

One  direction  which  appears  only  theoretically  possible  is  the  ex- 
ploitation in  governmental  affairs  and  international  relations  of  some  of 
the  implications  of  this  theory.  I  do  not  regard  this  as  likely  in  the  near 

I  suspect  that  the  discovery  and  development  of  a  contextual  basis 
for  this  theory  in  some  form  of  existential  philosophy  will  continue.  The 
general  orientation  of  philosophical  phenomenology  is  also  likely  to 
continue  to  have  its  influence  in  this  respect.  These  are  some  of  the 
potentialities  for  future  development — rather  grandiose,  to  be  sure — 
which  I  see.  The  extent  to  which  any  of  them  will  organically  grow  is 
a  matter  which  demands  a  gift  of  prophecy  I  do  not  have. 

Immediate  strategy  of  development.  To  return,  in  closing,  to  the 
much  more  immediate  issues  facing  us  in  the  systematic  development  of 
the  theory,  I  see  several  problems  which  have  very  high  priority  if  our 
general  systematic  thinking  is  to  have  a  healthy  development.  I  will  list 
these  problems  and  tasks,  but  the  order  of  listing  has  no  significance, 
since  I  cannot  determine  the  priority. 

L  We  are  urgently  in  need  of  new  and  more  ingenious  tools  of 
measurement.  Stephenson's  Q  technique  [81]  has  been  most  helpful  and 
Osgood's  method  for  quantifying  semantic  space  [51]  also  seems  promis- 
ing. But  most  urgently  needed  of  all  is  a  method  whereby  we  might 
give  operational  definition  to  the  construct  experience  in  our  theory,  so 

Therapy,  Personality,  and  Interpersonal  Relationships  251 

that  discrepancies  between  self-concept  and  experience,  awareness  and 
experience,  etc.,  might  be  measured.  This  would  permit  the  testing  of 
some  of  the  most  crucial  hypotheses  of  the  theoretical  system.  To  be  sure, 
some  attempts  have  been  made  to  approach  such  an  operational  defini- 
tion, but  the  instrumentation  is  exceedingly  cumbersome  and  admittedly 

2.  An  increased  amount  of  experience  with  individuals  classed  as 
psychotic,  and  the  testing  of  a  variety  of  the  theoretical  hypotheses  in 
therapeutic  work  with  this  group  and  in  research  with  psychotics  as  sub- 
jects, would  round  out  and  enrich  our  systematic  thinking  in  an  area 
in  which  it  is  at  present  inadequate.  It  would  provide  the  type  of  ex- 
treme reality  test  which  is  most  helpful  in  the  confirmation,  modification, 
or  disproof  of  a  theoretical  system.  There  would  seem  to  be  no  barriers 
except  practical  ones  to  such  a  development. 

3.  An  increased  amount  of  experience  and  careful  studies  of  hy- 
potheses developed  from  the  theory  are  needed  in  the  area  of  group 
relationships.  Hypotheses  regarding  leadership,  facilitation  of  learning, 
and  reduction  of  social  conflict  seem  particularly  fruitful  to  study.  Here 
again,  the  test  of  the  theory  at  one  of  its  deduced  extremes  would  be  most 
helpful  in  confirming  or  revising  its  core. 

4.  Still  another  urgent  need — no  doubt  quite  evident  to  readers  of 
this  presentation — is  the  translation  of  the  present  theory  into  terms 
which  meet  the  rigorous  requirements  of  the  logic  of  science.  Although 
progress  in  this  direction  has  been  made  there  is  still  a  woefully  long 
distance  to   go.   Such  a  development.,   carried  through  by  competent 
persons,  would  greatly  sharpen  the  deductive  hypotheses  which  might 
be  drawn  from  the  system,  and  hence  provide  more  crucial  tests  of  it. 

5.  The  final  need  I  wish  to  mention  may  seem  to  some  very  con- 
tradictory to  the  one  just  voiced.   Personally  I   see  it  as  a  possible 
evolutionary  step,  not  as  a  contradictory  one.  I  see  a  great  need  for 
creative  thinking  and  theorizing  in  regard  to  the  methods  of  social 
science.  There  is  a  rather  widespread  feeling  in  our  group  that  the  logical 
positivism  in  which  we  were  professionally  reared  is  not  necessarily  the 
final  philosophical  word  in  an  area  in  which  the  phenomenon  of  sub- 
jectivity plays  such  a  vital  and  central  part.  Have  we  evolved  the  optimal 
method  for  approximating  the  truth  in  this  area?  Is  there  some  view, 
possibly  developing  out  of  an  existentialist  orientation,   which  might 
preserve  the  values  of  logical  positivism  and  the  scientific  advances  which 
it  has  helped  to  foster  and  yet  find  more  room  for  the  existing  sub- 
jective person  who  is  at  the  heart  and  base  even  of  our  system  of  science? 
This  is  a  highly  speculative  dream  of  an  intangible  goal,  but  I  believe 
that  many  of  us  have  a  readiness  to  respond  to  the  person  or  persons  who 
can,  evolve  a  tentative  answer  to  the  riddle. 

252  CARL    R.    ROGERS 


I  find  myself  somewhat  appalled  at  the  length  and  scope  of  the  ma- 
terial which  has  been  presented.  I  suspect  the  reader  shares  this  feeling. 
I  can  only  say,  somewhat  apologetically,  that  I  had  not  fully  recognized 
the  ramifying  pervasiveness  of  our  theoretical  thinking  until  I  endeavored 
to  bring  it  all  under  one  verbal  roof.  If  many  of  the  outlying  structures 
appear  to  the  reader  flimsy  or  unfit  for  occupancy,  I  hope  that  he  will 
find  the  central  foundation,  the  theory  of  therapy,  more  solid.  If  to 
some  degree  this  formulation  bestirs  individuals  to  more  activity  in  re- 
search designed  to  prove  or  disprove  these  hypotheses,  or  to  more 
activity  in  building  a  better,  more  rigorous,  more  integrated  theory,  then 
the  group  which  is  collectively  responsible  for  the  foregoing  theories  will 
be  fully  satisfied, 


1.  Anderson,  R.  An  investigation  of  the  relationship  between  verbal  and 
physiological  behavior  during  client-centered  therapy.  Unpublished  doctoral 
dissertation.,  Univer.  of  Chicago,  1954. 

2.  Angyal,   A.   Foundations  for  a  science   of  personality.   New  York: 
Commonwealth  Fund,  1941. 

3.  Asch,  M.  J.  Nondirective  teaching  in  psychology:    an  experimental 
study.  Psychol  Monogr.,  1951,  65,  No.  4. 

4.  Assum,  A.  L.,  &  Levy,  S.  J.  Analysis  of  a  non-directive  case  with 
followup  interview.  /.  abnorm.  soc.  Psychol.,  1948,  43,  78-89. 

5.  Axline,  Virginia  M.  Play  therapy  and  race  conflict  in  young  children. 
/.  abnorm.  soc.  Psychol,  1948,  43,  300-310. 

6.  Butler,   J.   M.   Rank  pattern  analysis  of  counseling  protocols.   Un- 
published manuscript  (mimeo.).  Univer.  Chicago  Counseling  Center,  1954. 

7.  Butler,  J.  M.,  &  Haigh,  G.  V.  Changes  in  the  relation  between  self- 
concepts   and  ideal  concepts  consequent  upon  client-centered  counseling. 
In  [70,  chap.  4]. 

8.  Carr,  A.  C.  Evaluation  of  nine  psychotherapy  cases  by  the  Rorschach. 
/.  consult.  Psychol,  1949,  13  (3),  196-205. 

9.  Cartwright,  D.   Self-consistency  as  a  factor  in  affecting  immediate 
recall    Unpublished   manuscript    (mimeo.).    Univer.    Chicago    Counseling 
Center,  1955. 

10.  Chodorkoff,  B.  Self-perception,  perceptual  defense,  and  adjustment. 
/.  abnorm.  soc.  Psychol,  1954,  49  (4),  508-512. 

11.  Coch,  L.,  &  French,  J.  R.  P.,  Jr.  Overcoming  resistance  to  change, 
Human  Relat.,  1948,  1,  512-532. 

12.  Gofer,    C.    N.,    &    Chance,    J.    The    discomfort-relief    quotient    in 
published   cases  of  counseling  and  psychotherapy.   /.   Psychol,   1950,   29, 

Therapy,  Personality,  and  Interpersonal  Relationships  253 

13.  Cowen,  E.  L.,  &  Combs,  A.  W.  Followup  study  of  32  cases  treated 
by  nondirective  psychotherapy.  /.  abnorm.  soc.  PsychoL,  1950,  45,  232-258. 

14.  Diller,  L.  Conscious  and  unconscious  self -attitudes  after  success  and 
failure.  /.  Pers.,  1954,  23,  1-12. 

15.  Dymond,  Rosalind  F.  Adjustment  changes  over  therapy  from  self- 
sorts.  In  [70,,  chap.  5]. 

16.  Dymond,    Rosalind    F.    Adjustment    changes    over    therapy    from 
Thematic  Apperception  Test  ratings.  In  [70,  chap.  8]. 

17.  Faw,  V.  A  psychotherapeutic  method  of  teaching  psychology.  Amer. 
Psychologist,  1949,  4,  104-109. 

18.  Faw,    V.    Evaluation    of   student-centered    teaching.    Unpublished 
manuscript,  1954. 

19.  Fiedler,   F.    E.   A   comparative   investigation   of   early   therapeutic 
relationships  created  by  experts  and  non-experts  of  the  psychoanalytic,  non- 
directive  and  Adlerian  schools.  Unpublished  doctoral  dissertation,  Univer. 
of  Chicago,  1949. 

20.  Fiedler,  F.  E.  A  comparison  of  therapeutic  relationships  in  psycho- 
analytic, non-directive  and  Adlerian  therapy.  /.  consult.  PsychoL,  1950,  14, 

21.  Gallagher,  J.  J.  The  problem  of  escaping  clients  in  non-directive 
counseling.  In  W.  U.  Snyder  (Ed.),  Group  report  of  a  program  of  research 
In    psychotherapy.    Psychotherapy    Research    Group,    Pennsylvania    State 
Univer.,  1953.  Pp.  21-38. 

22.  Goldiamond,   I.   On   the  asynchrony  between  responses  in  a  per- 
ceptual experiment.  Unpublished  doctoral  dissertation,  Univer.  of  Chicago, 

23.  Gordon,  T.  Group-centered  leadership.  Boston:   Houghton  Mifflin, 

24.  Gordon,  T.  Group-centered  leadership  and  administration.  In  C.  R. 
Rogers,  Client- centered  therapy.  Boston:  Houghton  Mifflin,  1951.  Chap.  8. 

25.  Gordon,  T.,  &  Cartwright,  D.  The  effects  of  psychotherapy  upon 
certain  attitudes  toward  others.  In  [70,  chap.  1 1]. 

26.  Gross,   L.   An   experimental   study   of   the  validity  of  the  non-di- 
rective method  of  teaching.  /.  PsychoL,  1948,  26,  243-248. 

27.  Grummon,  D.  L.  The  use  of  language  categories  as  a  method  for 
study  of  personality  in  psychotherapy.  Unpublished  doctoral  dissertation, 
Univer.  of  Chicago,  1951. 

28.  Grummon,   D.   L.,   &  John,   Eve   S.   Changes   over  client- centered 
therapy  evaluated  on  psychoanalytically  based  Thematic  Apperception  Test 
scales.  In  [70,  chap.  11]. 

29.  Haigh,  G.  V.  Defensive  behavior  in  client-centered  therapy.  /.  con- 
sult. PsychoL,  1949,  13  (3),  181-189. 

30.  Haimowitz,   Natalie  Reader,  &  Morris,  L.   Personality  changes  in 
client-centered  therapy.  In  W.  Wolff  (Ed.),  Success  in  psychotherapy.  New 
York:  Grime  &  Stratton,  1952.  Chap.  3. 

31.  Haimowitz,  Natalie  Reader.  An  investigation  into  some  personality 
changes  occurring  in  individuals  undergoing  client-centered  therapy.  Un- 
published doctoral  dissertation,  Univer.  of  Chicago,  1948. 

254  CARL    R.    ROGERS 

32.  Hanlon,  T.  E.,  Hofstaetter,  P.  R.,  &  O'Connor,  J.  P.  Congruence  of 
self  and  ideal  self  in  relation  to  personality  adjustment.  /.  consult.  Psychol., 
1954,  18  (3) ,215-218. 

33.  Hartley,    Margaret.    Changes   in   the   self-concept   during   psycho- 
therapy. Unpublished  doctoral  dissertation,  Univer.  of  Chicago,  1951. 

34.  Hilgard,  E.  R.  Human  motives  and  the  concept  of  the  self.  Amer. 
Psychologist,  1949,  4,  374-382. 

35.  Hoffman,  A.  E.  A  study  of  reported  behavior  changes  in  counseling. 
/.  consult.  Psychol,  1949,  13,  190-195. 

36.  Hogan,  R.  The  development  of  a  measure  of  client  defensiveness 
in  the  counseling  relationship.  Unpublished  doctoral  dissertation,  Univer.  of 
Chicago,  1948. 

37.  John,  Eve  S.  Mental  health  and  the  principle  of  least  effort.  Un- 
published doctoral  dissertation,  Univer.  of  Chicago,  1953. 

38.  Jonietz,  Alice.  A  study  of  phenomenological  changes  in  perception 
after  psychotherapy  as  exhibited  in  the  content  of  Rorschach  percepts.  Un- 
published doctoral  dissertation,  Univer.  of  Chicago,  1950. 

39.  Kauffman,  P.  E.,  &  Raimy,  V.  C.  Two  methods  of  assessing  thera- 
peutic progress.  /.  abnorm.  soc.  Psychol.,  1949,  44,  379-385. 

40.  Kelley,  E.  G.  Education  in  communication.  ETC,  Summer,  1955, 

12,  248-256. 

41.  Kessler,  Carol.  Semantics  and  non-directive  counseling.  Unpublished 
master's  thesis,  Univer.  of  Chicago ,  1947. 

42.  Kirtner,  W.  L.  Success  and  failure  in  client-centered  therapy  as  a 
function  of  personality  variables.  Unpublished  master's  thesis,  Univer.  of 
Chicago,  1955. 

43.  Lecky,    P.   Self -consistency:    a   theory   of  personality.    New   York: 
Island  Press,  1945. 

44.  Lipkin,  S.  Clients'  feelings  and  attitudes  in  relation  to  the  outcome 
of  client-centered  therapy.  Psychol.  Monogr.,  1954,  68,  No.  1   (Whole  No. 

45.  Maslow,  A.   H.   Motivation   and  personality.   New  York:    Harper, 

46.  McCleary,  R.  A.,  &  Lazarus,  R.  S.  Autonomic  discrimination  with- 
out awareness.  /.  Pers.,  1949,  18,  171-1/9. 

47.  Mitchell,  F.  H.  A  test  of  certain  semantic  hypotheses  by  application 
to  client-centered  counseling  cases:  intensionality-extensionality  of  clients  in 
therapy.  Unpublished  doctoral  dissertation,  Univer.  of  Chicago,  1951. 

48.  Mosak,  H.  Evaluation  in  psychotherapy:   a  study  of  some  current 
measures.  Unpublished  doctoral  dissertation,  Univer.  of  Chicago,  1950. 

49.  Muench,  G.  A.  An  evaluation  of  non-directive  psychotherapy  by 
means  of  the  Rorschach  and  other  tests.  Appl.  Psychol.  Monogr.,  1947,  No. 

13,  1-163. 

50.  Nunnally,  J.  C.  An  investigation  of  some  propositions  of  self-con- 
ception: the  case  of  Miss  Sun.  /.  abnorm,  soc.  Psychol. ,  1955,  50,  87-92. 

51.  Osgood,  C.  E.  The  nature  and  measurement  of  meaning.  Psychol. 
Bull,  1954,  49,  197-237. 

Therapy,  Personality,  and  Interpersonal  Relationships  255 

52.  Quinn,  R.  D.  Psychotherapists'  expressions  as  an  index  to  the  quality 
of  early  therapeutic  relationships  established  by  representatives  of  the  non- 
directive,  Adlerian,  and  psychoanalytic  schools.  Unpublished  doctoral  dis- 
sertation, Univer.  of  Chicago,  1950. 

53.  Radke,  Marian,  &  Klisurich,  Dayna.  Experiments  in  changing  food 
habits.  /.  Amer.  dietetic  Ass.,  23,  403-409. 

54.  Raimy,  V.  C.  The  self -concept  as  a  factor  in  counseling  and  per- 
sonality organization.  Unpublished  doctoral  dissertation.,  Ohio  State  Univer., 

55.  Raimy,  V.   C.   Self  reference  in  counseling  interviews.  /.   consult. 
PsychoL,  1948,  12,  153-163. 

56.  Raskin,  N.  J.  An  objective  study  of  the  locus  of  evaluation  factor 
in  psychotherapy.  Unpublished  doctoral  dissertation,  Univer.  of  Chicago, 

57.  Roethlisberger,  F.  J.,  &  Dickson,  W.  J.  Management  and  the  worker. 
Cambridge,  Mass.:  Harvard  Univer.  Press,  1939. 

58.  Rogers,   C.   R.    Clinical  treatment   of  the  problem   child.   Boston: 
Houghton  Miffin,  1939. 

59.  Rogers,   C.  R.   Counseling  and  psychotherapy.  Boston:    Houghton 
Mifflin,  1942. 

60.  Rogers,  C.  R.  Some  observations  on  the  organization  of  personality. 
Amer.  Psychologist,  1947,  2,  358-368. 

61.  Rogers,    C.   R.   Dealing  with   social  tensions.   New  York:    Hinds, 
Hayden  and  Eldredge,  1948.  30  pp. 

62.  Rogers,  C.  R.  Client-centered  therapy.  Boston:   Houghton  Mifflin, 

63.  Rogers,   C.   R.    Communication:    its  blocking  and  its  facilitation, 
ETC,  Winter,  1952,  9,  83-88. 

64.  Rogers,    C.   R.   A   concept   of  the   fully  functioning  person.   Un- 
published manuscript  (mimeo.).  Univer.  Chicago  Counseling  Center,  1953. 

65.  Rogers,  C.  R.  The  implications  of  client-centered  therapy  for  family 
life.  Paper  given  to  Chicago  chapter  of  Int.  Soc.  Gen.  Semantics,  April, 

66.  Rogers,   C.   R.   Persons  or  science:    a  philosophical  question.  Am. 
Psychologist,    1955,  10,  267-278;  also  published  in  Cross  Currents,  Summer, 
1953,  3,  289-306. 

67.  Rogers,  C.  R.  The  case  of  Mrs.  Oak:   a  research  analysis.  In  [703 
chap.  15]. 

68.  Rogers,  C.  R.  Changes  in  the  maturity  of  behavior  as  related  to 
therapy.  In  [70,  chap.  13]. 

69.  Rogers,  C.  R.  This  is  me;  the  development  of  my  professional  think- 
ing and  my  personal  philosophy.  Paper  given  at  Brandeis  Univer.,  Nov., 

70.  Rogers,  C.  R.,  &  Dymond,  R.  F.   (Eds.)   Psychotherapy  and  per- 
sonality change.  Chicago:  Univer.  Chicago  Press,  1954. 

256  CARL   R.    ROGERS 

71.  Rogers,  Natalie.  Changes  in  self-concept  in  the  case  of  Mrs.  Ett. 
Personal  Counselor,  1947,  2,  278-291. 

72.  Rogers,   Natalie.   Measuring  psychological  tension  in  non-directive 
counseling.  Personal  Counselor,  1948,  3,  237-264. 

73.  RudikofT,  Esselyn  C.  A  comparative  study  of  the  changes  in  the  con- 
cept of  the  self,  the  ordinary  person,  and  the  ideal  in  eight  cases.  In  [70, 
chap.  11]. 

74.  Schwebel,  M.,  &  Asch,  M.  J.  Research  possibilities  in  non-directive 
teaching.  /.  educ.  PsychoL,  1948,  39,  359-369. 

75.  Seeman,  J.   Counselor  judgments  of  therapeutic  process  and  out- 
come. In  [70,  chap.  11]. 

76.  Seeman,    J.   A   study   of   the   process   of   non-directive   therapy.    /. 
consult.  PsychoL,  1949,  13,  157-168. 

77.  Seeman,  J.,  &  Raskin,  N.  J.  Research  perspectives  in  client-centered 
therapy.  In  O.  H.  Mowrer  (Ed.),  Psychotherapy:  theory  and  research.  New 
York:  Ronald,  1953. 

78.  Sheerer,  Elizabeth  T.  The  relationship  between  acceptance  of  self 
and  acceptance  of  others.  /.  consult.  PsychoL,  1949,  13  (3),  169-175. 

79.  Snyder,  W.  U.  An  investigation  of  the  nature  of  non-directive  psy- 
chotherapy. /.  genet.  PsychoL,  1945,  33,  193-223. 

80.  Standal,  S.  The  need  for  positive  regard:   a  contribution  to  client- 
centered   theory.    Unpublished   doctoral   dissertation,    Univer.    of   Chicago, 

81.  Stephenson,  W.  The  study  of  behavior:  Q-technique  and  its  method- 
ology. Chicago:  Univer.  of  Chicago  Press,  1953. 

82.  Stock,  Dorothy.   The   self  concept  and   feelings  toward  others.  /. 
consult.  PsychoL,  1949,  13  (3),  176-180. 

83.  Strom,  K.  A  re-study  of  William  U.  Snyder's  "An  investigation  of 
the  nature  of  non-directive  psychotherapy."  Unpublished  master's   thesis, 
Univer.  of  Chicago,  1948. 

84.  Thetford,  W.  N.  An  objective  measure  of  frustration  tolerance  in 
evaluating  psychotherapy.   In  W.  Wolff   (Ed.),  Success  in  psychotherapy, 
New  York:  Grune  &  Stratton,  1952.  Chap.  2. 

85.  Vargas,  M.  Changes  in  self -awareness  during  client-centered  therapy. 
In  [70,  chap.  10]. 

86.  Zimmerman,  J.  Modification  of  the  discomfort-relief  quotient  as  a 
measure  of  progress  in  counseling.  Unpublished  master's  thesis,  Univer.  of 
Chicago,  1950. 



Laboratory  of  Personality 
Assessment  and  Group  Be- 
havior,  University  of  Illinois 

Definition  of  the  Approach  {1} 257 

Personality  Research  in  Relation  to  the  Two  Basic  Scientific  Methods  {1  +}  .  261 

The  Logic  of  Factor  Analytic  Experiment  {4-6} 265 

The  Conceptual  Status  and  Interpretation  of  Factors  {3-5+} 270 

Classification  of  Factor  Phenomena  by  Modality,  Data,  and  Order  {3  +,  2}  .  273 

The  Present  Status  of  Findings  {8,  9} 278 

Complex  Function  and  Configural,  Type  Prediction  from  Source  Traits  {2+}  284 
The  Evidence  for  Motivational  and  Dynamic  Factors;  Ergs  and  Sentiments 

{2 +,9} 290 

The  Dynamic  Calculus  of  Ergic  Strengths,  the  Self  Sentiment,  Conflict,  and 

Integration  {2+} 296 

Summary  and  Systematic  Analysis  of  the  Present  Formulation  {2-12}  .     .     .     303 

Appendix:  The  Concept  of  Variable  Density  and  Factor  Order 318 

References 322 


The  maturity  of  theoretical  developments  may  be  tested  by  two 
touchstones.  First,  a  scientific  system  is  generally  more  mature  when  its 
concepts  arise  from  specially  developed  operations  and  techniques  other 
than  those  available  to  everyday  observation  and  to  the  layman.  Sec- 
ondly, theory  is  more  mature  if  we  can  point  to  ensuing  predictive  and 
controlling  powers  which  are  real  enough  to  have  led  to  potent  tech- 
nologies, recognizable  in  specially  developed  social  institutions, 

By  these  touchstones,  "personality  theory55  ranges  more  widely  in  de- 
velopmental level  than  do  most  other  areas  of  psychological  theory,  pre- 
senting examples  from  rarefied  heights  and  from  degraded  depths  of 
scientific  acceptability  and  status.  On  the  one  hand,  there  is  voluminous 
theory — principally  in  the  clinical  area — based  on  no  better  methods  of 


258  RAYMOND    B.    GATTELL 

observation  than  have  been  available  for  centuries,  entangled  in  verbal 
stereotypes  that  are  almost  certainly  false  or  purely  local  in  reference, 
intuitive  in  observation,  inexplicit  as  to  assumptions,  and  in  general,  not 
precisely,  operationally  based  or  confirmable.  From  this  level  of  scientific 
poverty  it  rises,  on  the  other  hand,  to  rational,  objective,  quantitative, 
and  intricately  developed  concepts  which  can  truly  be  said  to  surpass,  in 
both  complexity  of  testable  theory  and  effectiveness  of  technological  re- 
sults, such  neighboring  fields  as,  say,  learning  theory  and  group  dynamics. 
The  present  essay  is  concerned  exclusively  with  the  kind  of  person- 
ality theory  which  has  developed  out  of  quantitative  and  objective  meth- 
ods, whether  that  is  based  upon  clinical,  abnormal  data,  upon  social  or 
educational  fields  of  observation,  or  upon  laboratory  and  physiological 
study  of  the  normal  individual.  It  is  also  demarcated  from  neighboring 
developments  by  emphasis  on  multivariate  analytic  experiment  rather 
than  on  manipulative  univariate  experiment.  This  distinction  will  be 
drawn  more  clearly  in  a  moment;  at  the  outset,  let  it  be  said  that  in 
intention  researchers  in  the  present  area  are  aiming  at  an  experimentally 
based  personality  theory  and  that  the  emphasis  which  has  developed  on 
multivariate  rather  than  on  the  traditional,  controlled,  univariate  ex- 
perimental method  is  considered  only  an  intelligent  strategic  adaptation 
to  the  needs  of  personality  investigation  at  its  present  stage.  It  is  con- 
tended that  much  effort  has  been  relatively  wasted  in  unimaginative 
application  of  classical  experimental  design  by  psychologists  of  impec- 
cable scientific  aspirations,  who  have  failed  to  perceive  that  in  psychology 
(as  distinct  from  the  physical  sciences)   we  encounter  a  situation  and 
kind  of  data  to  which  classical  design  is  not  the  best  approach.  In  par- 
ticular, the  new  multivariate  experimenter  contends  that  classical  uni- 
variate experiment  has  insufficiently  realized:    (1)   that  in  psychology 
(compared  with  physics),  special  steps  must  be  taken  to  isolate  organi- 
cally unitary  and  unique  behavior  structures,  i.e.,  "significant  variables," 
before  univariate  experiment  can  be  strategically  applied,  ( 2 )  that  where 
so  many  variables  exist   (even  if  restricted  to  "significant"  ones)   the 
multivariate  approach  is  far  more  economical  and  powerful  in  mapping 
those  systematic  relations  among  variables  from  perception  of  which 
the  better-adapted  hypotheses  and  models  will  arise. 

This  contrast  with  univariate  experiment  is  not  the  whole  story 
regarding  the  character  of  the  multivariate  experimental  approach  con- 
sidered in  this  essay.  Indeed,  in  later  stages,  it  has  lost  some  of  the  char- 
acters which  initially  distinguished  it  from  the  univariate  experimental 
tradition;  notably,  it  has  begun  to  manipulate  variables  (thus  introduc- 
ing dependent-independent  variable  concepts),  although  in  a  framework 
of  simultaneous  operation  with  many  variables.  But  it  has  also  gathered 
new  characters  of  its  own  through  growing  into  fresh  branches  of  highly 

Personality  Theory  from  Quantitative  Research        259 

technical  specialized  experiment  and  conceptualization  which  have  no 
counterparts  in  univariate  experiment. 

In  conformity  with  the  plan  suggested  for  the  present  contributions 
on  systematic  viewpoints  we  shall  begin  with  background  factors  and 
orienting  attitudes.  Fortunately,  in  the  present  case  these  are  so  well 
known  that  only  a  very  brief  sketch  need  be  given.  The  main  historical 
roots  of  the  method  lie  directly  in  the  study  of  individual  differences, 
though  in  the  last  decade  it  has  emancipated  itself  completely  from  this 
restriction  and  has  been  largely  concerned  with  lifting  the  study  of  struc- 
ture and  process  to  new  technical  levels.  In  social  science  generally,  the 
multivariate  method  began  along  with  statistical  method,  or  at  least, 
with  the  second  covariational-analysis  phase  of  the  development  of 
statistics  by  Galton  and  Karl  Pearson.  In  psychology,  it  grew  to  a  lusty 
adolescence  in  the  study  of  individual  differences  in  ability  and  school 
achievement.  This  growth  began  with  Spearman's  and  Burt's  attempts 
fifty  years  ago  to  place  intelligence  testing  on  a  firm  basis  of  theory  and 
continued  through  Thurstone's  development  of  multifactor  analysis.  In 
some  isolated  backwaters  of  academic  teaching,  the  multivariate  ap- 
proach is  still  seen  in  these  terms  of  individual  differences,  of  nonmanip- 
ulative  experiment,  of  the  merely  economical  objective  of  finding  con- 
venient "dimensions,"  and  of  restriction  largely  to  educational  and 
cognitive  psychological  problems. 

Actually,  multivariate  methods,  of  which  factor  analysis  remains  the 
chief  development,  now  handle  far  more  issues  than  this  and  have  as 
much  to  contribute  in  personality,  learning,  and  motivation  study  as  in 
the  field  of  abilities.  They  offer  as  much  in  general  experimental  design 
as  in  the  psychometric  study  of  individual  differences.  It  is  a  truism  of 
scientific  history  that  classifications  which  appear  early  are  rarely  those 
which  are  ultimately  realized  through  the  logical,  inherent  characters 
of  the  methods  concerned.  The  approach  defined  here  may  be  seen 
historically  as  beginning  with  the  structural  and  taxonomic  problems  of 
classifying  abilities.  Yet  in  terms  of  its  inherent,  logical  nature  and  the 
real  applications  indicated  for  the  future,  multivariate  analysis  must  be 
seen  as  one  of  the  two  main  experimental  methods  available  in  science 
generally.  Incidentally  it  may  be  a  matter  of  justifiable  pride  to  psycholo- 
gists that  although  multivariate  analytical  methods  are  being  used, 
crescendo,  in  physiology,  medicine,  meteorology,  and  sociology,  they 
were  largely  developed  within  psychology  (as  the  univariate  methods 
were  within  the  older  physical  sciences) . 

Before  proceeding  to  bring  out  more  explicitly  the  procedures  and 
assumptions  of  the  multivariate  approach,  we  may  help  the  reader  to- 
ward perspective  by  giving  some  indication  of  substantive  content,  and 
also  of  the  relation  of  this  contribution  to  others  in  the  same  series. 


The  theories  developed  here  are  flanked,  on  the  motivational  side,  by 
Miller's  development  of  conflict  theory  (vol.  2)  and  Rapaport's  account 
of  psychoanalytic  structural  and  dynamic  concepts.  The  factor  analytic 
account  of  unitary  drives  needs  to  be  aligned  with  Morgan's  physiological 
picture  (vol.  1).  Our  discussion  of  factors  of  temperament  and  the 
methodology  of  psychological  genetics  should  be  brought  into  relation 
with  studies  of  human  heredity  in  various  fields  (e.g.,  Kallmann).  Our 
general  structural  theory  of  personality  has  affinities  to  and  differences 
from  the  stimulus-response  formulations  of  Hull  and  Spence  (vol.  2)  and 
the  more  clinical  analysis  by  Murray.  On  the  social  side,  our  mathe- 
matical models  for  attitudes  and  for  roles  can  be  related  to  the  contribu- 
tions of  Katz  and  Stotland  and  of  Newcomb,  respectively. 

As  these  relationships  are  studied,  it  will  become  evident  that  the 
present  approach  is  not  so  much  concerned  with  a  theory,  i.e.,  with  a 
particular  set  of  constructs  and  concepts  in,  for  example,  personality  and 
motivation,  as  with  many  possible  theories,  all  dependent  upon  the 
resolving  power  of  a  particular  methodological  approach.  Its  unity  is  not 
that  of  adherence  to  conceptual  beliefs  but  of  the  natural  integration 
which  exists  in  findings  from  a  particular  method  and  model,  flexibly 
applied  and  checked  against  other  methods  where  possible.  Nevertheless, 
we  admit  a  certain  attachment  to  the  theories  per  se,  and  certainly  we 
concentrate  on  the  theories  to  the  extent  that  this  essay  is  not  concerned 
with  all  derivatives  of  the  multivariate  method,  e.g.,  in  group  dynamics, 
culture-pattern  psychology,  physiology,  but  with  those  developed  in 
personality  and  motivation.  When  concentrating  on  the  theories,  how- 
ever, it  is  important  to  distinguish  them  from  superficially  similar  notions, 
often  with  similar  names  (e.g.,  Freud's  notion  of  ego  strength  or  Me- 
Dougall's  concept  of  the  self-regarding  sentiment) ,  which  do  not  arise 
from  this  mathematical  model  or  bear  the  hallmark  of  statistical  pre- 
cision in  measurement  which  the  present  concepts  always  imply. 

To  orient  the  reader  from  this  point  on,  the  author  should  state 
that  despite  his  intention  to  follow  the  excellent  editorial  outline  sug- 
gested for  all  contributions,  he  has  been  unable  entirely  to  adapt  the 
present  systematic  material  to  the  rubies  indicated.  Thus,  after  the  above 
statement  of  background  the  outline  proceeds  to  the  structure  of  the 
system,  particularly  the  systematic  independent,  intervening,  and  de- 
pendent variables.  This  sequence  is  ill-adapted  to  the  present  case  be- 
cause initially  there  are  no  dependent  and  independent  variables.  At 
least  in  the  factor  analytic  method  as  used  by  Spearman,  Burt,  and 
Thurstone  all  the  individual  difference  measurement  variables  stand  on 
an  equal  footing. 

Accordingly,  it  has  seemed  best  to  follow  an  order  of  exposition  which 
most  clearly  develops  an  understanding  of  the  dependence  of  ideas  upon 

Personality  Theory  from  Quantitative  Research        261 

procedures  (some  of  which  will  probably  not  be  initially  known  to  the 
reader)  and  then,  in  a  final  section,  explicitly  to  summarize  our  position 
in  terms  of  the  issues  raised  by  the  outline. 


After  attempting  to  demarcate  the  special  character  and  intention  of 
this  particular  contribution  to  personality  theory,  we  now  examine  more 
closely  its  chief  instrument — multivariate  analytical  experiment.  Since 
the  terms  univariate  and  multivariate  may  not  be  understood  in  the  same 
sense  by  all,  a  brief  comparative  analysis,  of  these  and  other  allegedly 
distinct  methods  used  in  personality  research,  is  necessary  before  empirical 
findings,  and  the  psychological  concepts  ensuing,  can  be  properly  focused. 

Actually,  it  is  common  to  hear  three  methods  mentioned  in  per- 
sonality research:  the  clinical  method,  the  controlled  experimental 
method,  and  the  multivariate  analytic  method — besides  special  emphases 
and  approaches  cutting  across  these,  such  as  the  anthropological,  the 
physiological,  etc.  In  the  controlled,  manipulative  classical  or  univariate 
experimental  method,  the  independent  variable  is  manipulated,  or 
allowed  to  alter,  while  all  other  variables  are  considered  to  be  controlled, 
except  for  the  changes  in  the  single  dependent  variable,  which  are 
recorded.  (Hence  univariate,  for  occasionally  the  independent  variable  is 
multiplied  to  two  or  three,  as  in  the  Fisherian  factorial  design.)  Ex- 
cept in  a  purely  positivistic  theoretical  framework,  the  empirical  in- 
dependent variable  is  understood  by  the  "classical"  experimenter  to 
represent  a  systematic  independent  variable — a  concept  or  construct 
which  he  has  postulated  to  be  so  represented.  But,  for  the  moment,  we 
shall  set  aside  what  the  experimenter  thinks  he  is  doing — for  this  can  be 
differently  conceived — and  ask  only  what  distinguishes  univariate  and 
multivariate  experiment  in  terms  of  what  the  experimenter  actually  does. 

Before  proceeding,  we  must  deny  the  third  approach,  the  "clinical 
method,"  any  status  as  a  fundamentally  distinct  method.  The  only 
logically  possible  treatments  of  relations  among  variables  are  in  pairs 
and  sequentially,  as  in  univariate  experiment,  or  in  large  numbers  and, 
usually,  without  knowledge  of  sequence,  as  in  multivariate  experiment. 
The  clinician  is  generally  a  multivariate  experimenter,  who  abstracts  laws 
and  concepts  from  observing  ("globally"  or  by  "gestalts"  as  he  might 
say)  simultaneous  changes  in  a  large  number  of  uncontrolled  variables. 
Fundamentally  he  does  exactly — or  perhaps  we  should  say  inexactly — 
what  the  factor  analyst  or  multivariate  experimenter  does,  but  he  does  it 
without  the  benefit  of  precise  instrumental  measurement  or  explicit 
correlational  procedures  (or  other  mathematical  treatment  of  functional 

262  RAYMOND    B.    CATTELL 

relations).  His  intuitions  about  functional  unities  are  thus  approxima- 
tions to  the  analyst's  independent  factors,  and  his  statements  about 
mutual  influences  of  factors  are  made  without  benefit  of  an  F  or  t  test. 
The  "clinical  method"  does  exist  and  function  usefully  as  a  rough 
"reconnaissance"  form  of  the  basic  multivariate  experimental  method. 
Claims  that  it  is  anything  other  than  this  confuse  the  clinical  method 
as  a  subdivision  of  therapy,  which  it  undoubtedly  is,  with  an  independent 
scientific  method,  which  it  undoubtedly  is  not.  For  such  claims  give 
merely  local  skills  and  methodological  accretions  grown  up  around 
clinical  practice  the  status  properly  due  only  to  a  fundamental  difference 
of  design. 

One  cannot  avoid  the  judgment  that  the  valuable  contribution  of 
clinical  practice  as  an  exploratory  method  has  lately  been  more  than 
offset  by  its  tendency  to  choke  the  growth  of  sound,  checkable  personality 
theory  in  a  rank  weedy  jungle  of  facile  verbal  concepts.  Where  quantita- 
tive and  computational  checks  are  not  possible — or,  at  least,  are  avoided 
by  the  formulation  of  "theories" — the  theoretical  field  becomes  a  mere 
playground  for  persons  of  high  fluency.  If  we  apply  the  test  of  maturity 
of  theory  suggested  above — the  production  of  an  effective  technology — 
then  the  few  existing  examinations,  notably  by  Kelley  and  Fiske  [67], 
Meehl  [76],  and  Eysenck  [50],  showing  that  clinical  psychology  is  in- 
distinguishably  above  chance  in  diagnosis  or  therapy,  leave  us  no  con- 
clusion but  that  purely  clinically  derived  theory  is  in  a  bad  way. 

Ironically,  however,  when  clinicians  or  others  have  tried  to  put  their 
house  in  order  and  to  extract  the  true  metal  of  science  from  the  ores 
in  which  clinical  data  are  richer  than  laboratory  data,  they  have  reverted 
to  classical  instead  of  that  multivariate  experiment  which  is  intrinsic  to 
"clinical  method"  and  the  potential  source  of  its  greatest  contribution. 
This  failure  is  rooted  partly  in  education — the  rarity  of  coordination  of 
mathematical  and  clinical  training — and  perhaps  partly  in  temperament. 
Lack  of  foresighted  handling  of  the  clinical  research  training  programs 
in  this  respect  is  likely  to  be  responsible  for  our  knowing  in  1970  laws 
which  we  might  have  known  and  applied  in  1960.  For,  before  controlled 
experiment  can  go  to  work  on  the  relation  of,  say,  superego  strength  to 
early  family  attitudes,  or  the  changes  of  "free"  and  "bound"  anxieties 
under  treatment  by  ataractic  drugs,  multivariate  research  must  first 
substantiate  the  existence  of  a  unitary  factor  of  superego  strength,  show 
tests  which  measure  it  with  defined  concept  validity  and  reliability,  and 
discover  whether  anxiety  in  fact  falls  into  one,  or  two,  or  more,  in- 
dependent sets  of  manifestations. 

The  emphasis  in  the  present  contribution  on  theory  derived  from 
multivariate  rather  than  univariate  quantitative  research  is,  in  summary, 

Personality  Theory  from  Quantitative  Research       263 

justified  both  by  the  historical  situation  and  by  the  intrinsic  logic  of 
method,  as  follows: 

First,  by  the  purely  sociohistorical  fact  mentioned  above,  that  the 
method  has  been  subject  to  gross  and  untimely  neglect  in  relation  to 
realistically  evaluated  potential  contribution.  It  has  been  neglected  be- 
cause, at  least  in  Germany  and  America,  those  interested  in  objective 
scientific  research  in  psychology  have  largely  been  conservatively  trained 
in  the  half-truth  that  in  psychology,  as  in  physics,  science  consists  of 
controlled  experiment. '  The  clinicians  who  had  the  courage  to  break 
away  from  this  tradition  realized  that  the  more  important  emotional 
situations  could  not  be  used  in  controlled  experiments  with  man.  A  rigid 
adherence  to  laboratory  experiment  would  lead  to  the  restriction  of  data 
to  such  specialized  but  "trivial"  fields  as  perception,  the  psychophysiology 
of  reflexes,  or  the  sense  organs,  or  to  experiments  on  the  emotions  of 
animals,  which  could  never  be  applied,  except  by  uncertain  analogy,  to 
the  personalities  of  human  beings. 

Secondly,  at  this  primitive  stage  of  personality  research  especially, 
the  multivariate  method  offers  a  swifter  and  surer  approach  to  the 
significant  variables  for  controlled  experimentation.  In  personality,  as  in 
psychology  and  the  life  sciences  as  a  whole,  the  investigator  has  an  in- 
finite array  of  variables  from  which  to  choose.  It  is  not  surprising — and 
is  perhaps  a  comment  on  our  ways  of  striving  for  originality — that  one 
and  the  same  empirical  (not  conceptual)  variable  rarely  gets  confirm- 
atory investigation  by  as  many  as  two  psychologists.  Apparently,  there 
are  at  least  as  many  variables  claimed  to  be  of  outstanding  significance  as 
there  are  psychologists. 

One  of  the  common  schemata  underlying  presentations  in  this  book 
is  the  statement  of  independent,  intervening,  and  dependent  variables  in 
each  field.  It  has  been  editorially  suggested  that  an  independent  (or 
dependent)  variable  should  be  further  considered  in  experimental, 
mathematical,  and  ideational  (systematic)  senses.  This  initial  clarification 
is  best  adapted  to  the  univariate  methodology  from  which  it  was  de- 
rived; in  the  multivariate  field  it  needs  further  structuring.  A  factor  is 
both  a  systematic,  conceptual  independent  (or  dependent)  variable  and 
an  intermediate  variable.  The  strict  multivariate  methodologist  is  un- 
likely, indeed,  to  introduce  any  conceptual  intermediate  variable  that 
is  not  first  revealed  as  a  unitary  factor.  However,  the  proof  that  a  uni- 
tary entity  exists,  and  that  it  is  therefore  profitable  to  begin  setting  up 
hypotheses  about  it,  as  a  unitary  concept,  may  occur  years  ahead  of  the 
confirmation  of  what  the  entity  is. 

The  contention  of  the  multivariate  analyst  is  that  too  many  psy- 
chologists have  immaturely  "jumped  the  gun"  by  imitating  the  univari- 

264  RAYMOND    B.    CATTELL 

ate  experiment  of  physics  in  psychology,  without  regard  to  the  different 
stages  and  natures  of  these  sciences.  Greater  shrewdness  might  have 
foretold  the  impotence  which  these  incontinent  procedures  have  demon- 
strated during  more  than  half  a  century.  It  would  have— and  now  em- 
phatically has — indicated  that  a  better  strategy  is  to  reduce  the  chaos 
of  infinite  possible  variables  to  more  tractable  and  significant  numbers 
and  natures  by  factor  analysis  before  much  hypothesis  formation  and 
manipulative  experiment  begins.  A  decade  devoted  largely  to  systematic, 
cooperative  studies  of  this  kind  would  not  now  be  out  of  place  in  per- 
sonality study,  or  indeed  in  learning,  physiological,  and  social  psychology. 
The  third  and  last  relation  to  be  emphasized  between  multivariate 
and  univariate  methods  justifies  greater  resort  to  the  former  not  only 
because  of  aptness  to  the  present  developmental  phase,  but  generally.  This 
is  the  argument  from  research  economy  and  certainty  of  inference.  It 
springs  from  three  sources : 

1.  One  multivariate  research  with,  say,  30  variables  yields  evidence 
on    (30X29)/2  =  435  relationships,  with  only  fifteen  times  the  ex- 
perimental work  required  in  one  univariate  experiment.  Consequently  it 
achieves  the  results  of  435   univariate  experiments  with   about  one- 
thirtieth  of  the  expenditure. 

2.  The  relationships  are  determined  under  conditions  in  which  all 
variables  are  allowed  to  vary  over  their  full  range  together.  Consequently 
one  does  not  have  the  uncertainty,  which  occurs  in  trying  to  make  in- 
ferences from  many  univariate  experiments,  as  to  possible  interaction 
effects  lost  through  the  controlled  situation  or  as  to  corrections  necessary 
in  integration  because  the  different  univariate  relations  have  been  found 
on  diverse  samples. 

3.  The  hypothesis  being  tested  is  made  more  determinate  through 
being  represented  by  a  factor  measurement  based  on  several  empirical 
independent  variables  instead  of  one  only.  For  example,  an  investigator 
may  set  out  to  test  the  hypothesis  that  rigidity  is  related  to  rate  of  con- 
ditioning and  state  that  operation  X  defines  operationally  his  hypothesis 
or  concept  of  the  nature  of  rigidity,  i.e.,  X  is  the  empirical  independent 
variable  defining  his  systematic  independent  variable,  rigidity.  But  fac- 
tor ^  analysis  might  show,  as  it  frequently  has,  that  only  one-third  of  the 
variance  of  X  is  accounted  for  by  a  rigidity  factor  and  that  the  rest  is 
equally  determined  by  two  other  factors,  say,  intelligence  and  fatigue. 
It  may  take  several  blind  conceptual-trial-and-error  studies  before  the 
univariate  experimenter  hits  on  a  better  variable  to  represent  rigidity 
(as  judged  by  more  consistent  or  positive  results),  and  during  that  time 
his^  conclusions  could  just  as  well  be  statements  about  relations  of  con- 
ditioning to  intelligence  or  fatigue  as  to  rigidity.  A  factorist  would  first 
determine  the  factor  structure  and  then  tie  the  factor  down  by  several 

Personality  Theory  from  Quantitative  Research        265 

operational  representatives.  For  it  is  rare  to  find  a  complex  concept  that 
can  be  represented  by  a  single  operation,  and  it  is  still  more  rare  for  a 
univariate  experimenter  to  land  on  it  at  the  first  attempt. 

Against  these  substantial  advantages  two  shortcomings  can  be 
charged  to  the  multivariate  method,  ( 1 )  that  it  only  handles  linear  re- 
lationships, ( 2 )  that  it  omits  time  sequence  and  therefore  does  not  per- 
mit unambiguous  causal  inference.  (More  narrowly  stated,  it  deals  with 
response-response  relations  rather  than  stimulus-response.)  The  first  is 
true,  but  it  is  generally  desirable  to  observe  any  relation  in  the  approxi- 
mate, linear  form  before  proceeding  to  more  complex  functions.  How 
many  relations  do  we  yet  know  of  in  psychology  involving  a  law  that  is 
indubitably  different  from  one  of  simple  proportion?  And  are  not  most 
controlled  experiments  content  with  an  analysis  of  variance  significance 
test,  proving  nothing  at  all  about  the  form  of  the  relation?  As  to  the 
second,  it  rests  largely  on  lack  of  reading  in  multivariate  methods.  The 
condition-response  factor  design  [23,  42]  systematically  investigates  the 
relation  of  controlled  changes  of  stimulus  to  response;  P  technique  [35] 
and  incremental  R  technique  use  factor  analysis  over  time  intervals 
rather  than  in  instantaneous,  nonsequential  analysis. 

These  later,  more  developed  multivariate  designs  permit  causal  in- 
ference about  interaction  of  factors  to  be  drawn  from  the  same  experi- 
ment as  that  which  structures  the  variables  into  factors,  as  will  be  seen 
in  examples  in  the  following  sections.  They  retain,  however,  the  ad- 
vantage that  manipulative  control  of  most  variables  is  not  necessary,  as 
it  typically  has  to  be  in  most  univariate  experiment.  Instead  of  "isolating 
by  control,"  the  multivariate  experiment  allows  nature  to  vary  as  it  will 
(often  producing  effects  we  should  not  dare  to  duplicate  in  human  ex- 
periment) and  then  isolates  by  superior  statistical  analysis  what  cannot 
be  isolated  by  physical  manipulation.  For  example,  one  might  be  in- 
terested in  the  effects  of  Group  Morale  Factor  2  [41]  upon  individual 
responses  expressed  in  murder  rates.  Fortunately  we  are  spared  responsi- 
bility for  the  latter  because  we  do  not  know  how  to  manipulate  factor  2, 
but  we  can  accurately  measure  its  changes  and  investigate  the  relations 
accordingly.  The  wider  realm  of  multivariate  experimental  design  can 
be  read  about  elsewhere  [23,  28,  58,  90],  so  we  shall  now  confine  our- 
selves to  the  relevant  essentials  of  the  factor  analytic  model. 


We  need  not  deal  with  the  mathematics  and  the  computational  pro- 
cedures of  factor  analysis  here  [see  23,  58,  90],  but  its  logic  should  be 
briefly  stated.  Any  of  the  standard  factor  analytic  procedures  will  reduce 
the  variance  on  a  large  member  n  of  individual  variables  to  variance  on 

266  RAYMOND    B.    CATTELL 

a  small  number  of  common  factors  k  plus  variance  on  n  specific  factors. 
Thereafter  the  score  P  of  a  person  i  on  a  specific  variable  ;  can  be  es- 
timated by  the  specification  equation : 

Pji  =  SjiFu  +  SjtFzi  +  *  •  •  +  SjhFki  +  SjFj* 

where  the  /s  are  the  situational  indices  or  loadings,  obtained  by  factor- 
ing the  correlation  matrix  for  the  n  variables,  and  the  F's  are  the 
strengths  of  the  endowments  of  the  individual  i  in  the  various  factors. 
Factors  1  to  n  are  factors  common  to  this  and  other  performances, 
whereas  factor  ;  is  specific  to  this  particular  response.  The  factor  matrix, 
obtained  by  factor  analytic  procedures  from  the  correlation  matrix, 
gives  us  all  we  need  for  the  above  general  equation.  Each  row  of  the 
matrix  gives  the  set  of  /s  for  estimation  of  the  given  variable,  and  each 
column,  presenting  a  factor,  shows  which  variables  need  to  have  their 
weighted  scores  added  together  to  give  an  estimate  of  that  factor  for  any 

It  will  be  noted  in  passing  that  this  formulation  again  transcends,  or 
requires  a  new  view  of,  the  reduction  of  scientific  systems  to  independent 
and  dependent  variables  and  intermediate  variables  or  constructs.  For 
the  initial  variables  are  ( at  least  in  timeless,  instantaneous  factor  analysis ) 
both  the  independent  variables  from  which  the  construct — the  factor — is 
inferred,  and  the  dependent  variables  predicted  from  these  intermediate 
variables,  in  the  specification  equation. 

The  majority  of  factor  analytic  researches  are  not  carried  out  with 
the  object  of  proceeding  to  actual  specification  equation  computations 
but  rather  with  the  general  scientific  aim  of  determining  the  number  and 
nature  of  the  psychological  factors  at  work  in  a  given  phenomenal  area. 
At  this  level,  issues  have  been  much  confused  by  difference  of  purpose 
between  mathematical  statisticians  and  psychological  researchers.  The 
mathematical  statisticians  are  content  if  they  can  find  a  reduced  number 
of  orthogonal  factors  which  will  reproduce  the  correlations,  and  if  pos- 
sible also  the  given  scores,  within  the  given  experiment.  But  psychologists 
are  concerned  to  know  that  they  have  found  the  correct  number  of 
factors  and  that  they  have  the  correct  nature  (pattern)  for  each  factor, 
in  terms  of  other  experiments  beyond  the  one  in  question,  i.e.,  in  terms  of 
general  scientific  concepts.  Consequently  psychologists  do  not  see  ad- 
vantage in  the  mathematical  neatness  of  orthogonality;  they  positively 
reject  it,  because  it  is  highly  probable  that  all  factors  in  the  same  universe 
have  interaction  and  are  likely  to  be  somewhat  correlated  among 

The  mathematician  knows  many — indeed  an  infinite  number — of 
combinations  of  numbers  and  natures  of  factors  that  will  reproduce  the 

Personality  Theory  from  Quantitative  Research       267 

given  variable  correlations  and  scores,  and  if  he  prefers  any  one,  it  will 
be  for  mathematical  neatness.  The  psychologist  wants  conditions  for  de- 
termining a  unique  solution,  i.e.,  a  fixed  number  of  factors  rotated  to 
one  fixed  position,  and  he  is  more  concerned  that  this  unique  interpre- 
tation fit  the  interpretation  of  other  experimental  matrices  than  that  it 
fit  certain  concepts  of  mere  mathematical  convenience  within  one  matrix. 
The  pursuit  of  this  latter  aim  is  tied  up  technically  with  development  of 
(1)  communality  estimation  theories,  (2)  the  invention  of  formulas  for 
standard  errors  for  factors  and  loadings,  and  (3)  the  determination  of 
unique  rotation  positions  by  simple  structure,  criterion  rotation,  and 
parallel  profiles  [1,7,  23,  29,  31,  44,  81,  90,  95],  In  these,  statistical  logic 
has  had  at  times  to  limp  along  with  the  help  of  a  crutch  derived  from 
empirical  generalizations;  but  as  of  1958,  the  major  problems  have  been 
overcome  just  sufficiently,  though  not  always  to  the  satisfaction  of  the 
theoretical  statistician.  That  is  to  say,  researchers  will  in  general  now 
agree  on  how  many  factors  there  are,  and  the  "payoff3  of  arranging 
findings  from  many  studies  side  by  side  shows  that  simple  structure  is 
capable  of  revealing  the  same  factor  patterns  from  different,  independent 
experimental  studies.  An  essential  part  of  this  completion  of  adequate 
techniques  has  been  the  development  of  factor-matching  indices,  such  as 
the  recent  formula  by  Gattell  and  Baggaley  [31],  which  permit  us  to 
give  fiducial  limits  to  the  goodness  of  a  given  matching  of  factors  from 
one  study  to  another.  With  improved  techniques  for  obtaining  unique 
resolution  into  factors,  and  improved  methods  of  checking  factors  from 
study  to  study,  it  has  been  possible  to  demonstrate  the  in  variance  of  10 
to  20  personality  and  ability  factors. 

The  logic  of  the  resolution  of  variance  on  many  variables  into  a  set 
of  unique  common  factors,  specifics,  and  error,  is  the  same  for  all  factor 
analytic  designs,  regardless  of  experimental  setting.  But  the  uses  of  factor 
analysis  in  different  contexts  of  stimulus,  response,  and  organism,  and  the 
scientific  meaning  of  the  factors  derived  therefrom,  fall  basically  into  six 
distinct  experimental  designs — actually  a  set  of  three  basic  designs,  each 
analyzable  in  two  different  ways  [22,  23].  The  three  basic  designs  arise 
from  the  nature  of  behavioral  measurement.  Any  behavioral  measure- 
ment is  defined  and  tagged  by  five  referents:  a  particular  organism 
making  the  response,  a  particular  stimulus  situation  in  which  the  re- 
sponse occurs,  a  particular  moment  in  time,  a  particular  point  in  space, 
and  a  particular  observer  [22].  Setting  aside  the  two  last  as  irrelevant  to 
the  basic  designs,  we  have  three  characteristics,  any  one  of  which  can 
be  repeated  many  times  to  create  the  series  of  entries  required  for  cor- 
relation purposes.  Thus  we  can  have  the  same  stimulus  situation  and 
class  of  response,  measured  at  the  same  moment  in  time  on  a  series  of 
different  organisms  of  the  same  class.  This  is  the  traditional  correlation 


procedure,  e.g.,  measuring  a  set  of  schoolboys  on  their  response  to  an 
intelligence  test  and  then  on  a  mathematics  test  and  correlating  the  two 
series.  When  carried  to  a  factor  analysis  it  is  called  R  technique.  Secondly, 
we  may  correlate  over  a  series  of  occasions  (moments  in  time)  instead 
of  a  series  of  persons,  taking  again  and  again  the  same  set  of  stimulus- 
response  (test  situation)  measures  upon  one  person.  This  is  called  P 
technique.  The  three  basic  designs,  or  experimental  possibilities  of  cor- 

FIG.  1.  The  covariation  chart.  From  [14]. 

relation,  are  called  P,  R,  and  T  techniques  and  are  shown,  with  their 
transposes,  in  the  Covariation  Chart  in  Fig.  1. 

The  meaning  of  the  various  possibilities  in  this  chart  the  reader  may 
work  out  for  himself,  or  consult  fuller  accounts  [14].  Until  the  gen- 
eralized statement  of  covariation  possibilities  was  published  [14]  in 
1946,  about  99  per  cent  of  all  correlation  and  factoring  had  been  R 
technique  and  the  rest  Q  and  P  techniques.  Since  1946,  most  of  the 
theoretical  possibilities  have  been  tried  in  practice  and  there  has  been 
much  wider  use  of  P  and  Q  techniques.  It  will  be  observed  that  each  of 
the  three  basic  designs,  R,  P,  and  T,  permits  a  transposed  factoring  of 
the  same  score  matrix,  namely,  Q,  O,  and  S,  respectively.  It  is  now 

Personality  Theory  from  Quantitative  Research       269 

generally  accepted  that  the  same  factors  are  obtainable  from  any  de- 
sign and  its  transpose,  so  that  the  decision  to  use  one  or  the  other  of 
the  pair  depends  upon  convenience.  For  example,  with  many  subjects 
and  few  tests,  R  technique  is  appropriate;  the  converse  suggests  the  Q- 
technique  transpose. 

Furthermore,  the  three  basic  designs  and  their  transposes  may  be 
modified  further  to  produce  several  other  useful  designs.  For  example, 
incremental  R  technique,  instead  of  factoring  an  absolute  score  for  each 
person,  can  enter  correlation  with  the  difference  of  score  for  each  person 
between  stimulus  occasions  one  and  two  [28,  102].  Moreover,  as  sug- 
gested above,  although  factor  analysis  grew  up  using  naturally  occurring, 
not  experimentally  created  variation,  nothing  intrinsic  to  multivariate  de- 
sign prevents  its  also  being  used  with  varying  stimuli  as  well  as  with  vary- 
ing responses.  What  has  been  called  the  condition  response  design  [23], 
which  randomizes  several  controlled,  varying  stimulus  conditions  with 
respect  to  one  another,  is  one  example  of  such  use.  Essentially,  it  factors 
stimuli  and  responses  together,  obtaining  at  once  the  unitary  patterns 
of  response  and  their  relations  to  stimulus  conditions. 

No  thorough  treatment  of  the  varied  possible  factor  analytic  experi- 
mental designs  is  possible  here.  Our  objective  in  glancing  over  them  is 
merely  to  point  out  that,  as  conceived  by  the  psychologist,  the  factor, 
or  "source  trait"  [14],  differs  from  a  mere  mathematical  factor  not  only 
by  reappearing  in  several  distinct  R-technique  studies,  as  already  men- 
tioned, but  also  by  its  capacity  to  reappear  as  the  same  pattern  in  these 
different  experimental  designs.  For  example,  a  factor  labeled  Surgency- 
and-Desurgency  has  been  found  in  R-technique  analysis,  in  terms  of  in- 
dividual differences,  loading  such  manifest  variables  as  cheerful,  im- 
pulsive, talkative,  unworried,  and  some  physiological  variables,  notably 
serum  cholinesterase  concentration.  When  the  same  variables  are  meas- 
ured from  day  to  day  on  a  single  individual  and  their  trends  are  inter- 
correlated,  the  factor  analysis  produces  an  intraindividual  pattern  of 
just  the  same  form.  That  is  to  say,  Surgency-Desurgency  is  a  unity  in 
terms  of  individual  differences  and  also  in  terms  of  function  fluctuation 
within  one  person. 

Finally,  it  should  be  recognized  that  a  more  generalized  factor 
analytic  model  does  not  preclude  nonparametric  variables  or  functional 
relations  of  factors  with  variables  more  complex  than  those  of  simple 
linearity,  as  shown  by  Coombs  and  Satter  [43]  and  discussed  more  fully 
elsewhere  [28].  Until  the  modified  models  are  developed  in  terms  of  com- 
putational analysis  it  will  still  be  necessary  to  find  the  factors  by  the 
present  linear  approximation,  operating,  therefore,  over  the  small  vari- 
ance ranges  where  the  approximation  better  holds.  But  once  the  source 
traits  are  recognized  and  measurable,  the  nonlinear  formulas  better  qx- 


pressing  their  relation  to  any  particular  criterion  can  be  more  accurately 
determined  in  the  usual  curve-fitting  way. 

As  shown  elsewhere  [28],  approaches  in  terms  of  patterns  and  types 
are  merely  the  obverse  of  the  factor  approach.  The  methods  suggested 
by  Horst,  Lubin,  Meehl,  McQuitty,  Ellson,  Gibson,  and  others  for 
finding  or  using  type,  pattern,  or  profile  functions  are  most  simply 
applied  to  factors  rather  than  single  variables.  Indeed  there  are  serious 
objections  to  applying  them  to  variables  [28].  A  whole  new  development 
of  theoretical  understanding  lies  open,  in  terms  of  pattern  emergent 
junctions  applied  to  -factor  profiles,  once  our  grasp  of  the  number  and 
nature  of  personality  factors  in  man  has  reached  acceptable  precision. 


The  notion  that  a  factor  is  a  single  unitary  influence  underlying  many 
manifestations  rests  on  the  logical  premise,  found  for  example  in  the 
writings  of  the  logician  J.  S.  Mill,  that  covariation  betokens  a  common 
cause  or  elements.  It  has  been  suggested,  however,  [14]  that  in  psy- 
chology we  should  not  conceive  a  unity  as  an  all-or-nothing  "reality"  but 
admit  degrees  of  efficacy  or  potency.  A  certain  unity  may  show  itself 
in  R-technique  studies  with  adults  but  not  with  children,  as  some  of  the 
primary  abilities  do.  Another  may  show  itself  in  all  R-technique  analyses 
but  not  in  P  technique,  as  general  body  size  and  general  intelligence  do, 
and  so  on.  In  general,  a  unitary  influence  is  capable  of  maintaining 
its  unity  only  through  certain  ranges  of  conditions.  Parenthetically,  a 
unity  which  shows  itself  in  P  technique  should  in  general  be  expected  to 
show  itself  in  R  technique,  but  not  vice  versa;  for  the  levels  of  variables 
as  caught  at  a  given  moment,  in  a  given  population  sample,  contain  both 
the  fixed  individual  differences  and  the  internal  fluctuation,  i.e.,  they 
represent  both  a  trait  and  a  state.  Recent  evidence  agrees  in  systematically 
turning  up  more  factors  for  the  same  set  of  variables  in  R  technique 
than  in  P  technique  [28]. 

At  this  point,  we  may  state  that  not  merely  will  a  unitary  influence 
show  itself  as  a  factor,  but  that  no  inference  about  the  existence  of 
unitary  influences  is  possible,  by  known  scientific  method,  except  through 
multivariate  analysis,  over  the  range  of  designs  here  listed.  The  existence 
of  a  unity  cannot  be  proved  by  intuitive  perception,  or  by  univariate  ex- 
periment, or  by  clinical  inference.  (For  the  last  is  but  an  approximation 
to  the  statistical  multivariate  analysis  procedures.)  However,  a  con- 
trolled experiment  may  be  part  of  the  proof  of  existence  of  a  unity, 
provided  that  the  experiment  has  multiple  dependent  variables.  For 
example,  Cureton  obtained  six  R-technique  factors  in  about  one  hundred 
physical-performance  variables,  and  labeled  one  of  these  factors  ucardio- 

Personality  Theory  -from  Quantitative  Research        271 

vascular  efficiency,"  i.e.,  capacity  to  bring  oxygen  to  the  tissues.  Later 
experiments  with  the  same  variables  in  a  "high  altitude"  oxygen  decom- 
pression chamber,  using  oxygen  pressure  as  the  single  independent 
variable,  showed  that  it  was  the  particular  pattern  of  variables  loaded  in 
this  factor,  and  no  others,  which  showed  deterioration  with  drop  in 
oxygen  tension.  Another  example  is  the  discovery  of  the  same  particular 
stress-response  pattern,  on  the  one  hand,  by  Selye's  fitting  together  of 
evidence  from  several  kinds  of  controlled  experiment  on  animals  and,  on 
the  other,  by  our  P-technique  factoring  of  the  same  variables  in  human 
beings  [27].  On  the  other  hand  the  checking  of  a  hypothesis  about 
unitariness  by  a  hypothetico-deductive  sequence  of  univariate  experi- 
ments is  less  satisfactory,  because  of  our  inability  confidently  to  integrate 
correlational  evidence  from  many  different  samples. 

Granted  that  a  unitary  pattern  is  established,  qua  pattern — and  this 
requires  statistical  checks,  such  as  the  salient  variable  similarity  index — 
the  next  step  is  its  interpretation,  as  a  cause  or  dimension,  or  at  least  the 
formulation  of  a  progressively  testable  hypothesis  about  it.  The  interpreta- 
tion of  a  factor  is  sometimes  made  by  considering  the  set  of  variables 
highly  loaded  in  it,  i.e.,  those  whose  variance  is  substantially  accounted 
for  by  the  factor,  and  seeking  to  abstract  some  quality,  content,  or 
principle  common  to  them  all.  This  is  all  right  as  far  as  it  goes,  but  the 
complete  procedure  requires  attention  not  only  to  the  variables  highly 
positively  and  negatively  loaded  but  also  to  those  with  essentially  zero 
loadings.  For  we  infer  the  nature  of  a  thing  not  only  from  what  happens 
when  it  is  present  but  also  from  what  happens  when  it  is  absent.  More- 
over, the  psychologist  needs  to  be  more  alert  than  he  generally  is  to  the 
possibility  that  a  variable  important  to  his  deductions  actually  was  ex- 
perimentally included  in  the  researches  he  is  surveying.  There  are  in- 
stances of  psychologists  forming  hypotheses  on  the  assumption  that  a 
given  variable  forms  no  part  of  the  factor  pattern  when  it  was,  indeed, 
never  included  in  the  correlation  matrix  and  so  could  not  possibly 
manifest  a  loading. 

The  process  of  deeper  interpretation  of  a  manifest  factor  pattern  in 
terms  of  a  source  trait  entity  consists  usually  of  a  hypothetico-deductive 
experimental  sequence.  Incidentally,  it  has  often  been  maintained,  even 
by  factor  analysts  [49],  that  the  multivariate  design  differs  from  the 
controlled  univariate  experiment  in  not  being  hypothetico-deductive.  On 
first  seeing  a  factor  loading  pattern,  meeting  the  conditions  of  simple 
structure,  the  experimenter  forms  a  hypothesis  about  the  nature  of  the 
source  trait.  From  this  he  deduces  that  a  previously  unused  variable 
A  should  be  more  highly  positively  loaded  than  anything  he  now  has 
in  the  matrix,  that  another,  B,  should  be  more  negatively  loaded,  and  that 
a  third,  C,  should  be  unaffected.  With  these  three  (or  more)  new 

272  RAYMOND    B.    CATTELL 

variables  he  reenters  experiment,  to  see  whether  his  deduction  is  con- 

Incidentally,  we  should  note  that  this  is  a  more  logically  exacting, 
and  frequently  a  more  statistically  exacting,  test  of  a  hypothesis  than 
the  mere  establishment  of  significant  difference  on  a  single  variable,  as  in 
univariate  experiment.  For  in  the  former  the  experimenter  predicts  that  a 
whole  pattern  of  variables  will  behave  in  a  certain  fashion,  whereas  the 
fact  that  a  single  variable  increases  or  decreases  as  predicted,  in  a  uni- 
variate experiment,  usually  leaves  inference  much  more  undetermined. 
However,  it  is  very  rarely  that  a  correct  hypothesis  for  a  factor  has  been 
reached  in  a  single  act  of  reasoning,  and  more  commonly  we  proceed 
through  a  spiral  of  hypotheses  and  experiments,  gradually  raising  the 
loadings  of  variables  toward  that  value  of  unity  (when  corrected  for 
attenuation)  which  permits  us  to  say  we  have  found  the  underlying 
variable  which  is  the  factor. 

In  this  connection  we  should  note  that  though  "factor"  and  "source 
trait"  are  often  used  as  synonyms,  yet  there  is  in  fact  a  conceptual 
duality.  On  the  one  hand,  we  have  the  factor  (not  necessarily  a  factor  in 
a  single  matrix)  which  is  strictly  a  factor  pattern  of  loadings,  as  inferred 
for  a  parent  population;  on  the  other,  we  have  the  concept  of  a  single 
underlying  "intermediate  variable"  [75]  which  causes  this  pattern.  The 
pattern  is  our  only  means  of  referring  to  the  source  trait,  of  recognizing 
and  defining  it.  (At  least,  unless  there  is  supplementary  controlled  ex- 
perimental evidence  as  mentioned  above.)  And  yet  we  know  that  this 
pattern  can  never  be  exactly  the  same  from  one  sample  to  another,  be- 
cause of  sampling  and  experimental  error;  or  from  one  population  to 
another,  because  of  systematic  influences;  or  from  one  technique  to 
another,  since,  for  example,  some  variables,  which  do  not  vary  from  per- 
son to  person,  can  fluctuate  in  P  technique,  over  time,  and  vice  versa. 
The  source  trait  is  the  entity,  whether  it  remains  abstractly  a  construct 
and  concept,  or  comes  to  be  representable  by  a  literal  variable  never  seen 
before;  whereas  the  factor  is  only  a  pattern  found  in  some  complex 
statistical  derivatives  called  loadings. 

The  identification  of  the  source  trait  from  the  pattern  can  always  be 
made  by  understanding  and  applying  the  statistical  and  other  laws  which 
produce  the  various  pattern  modifications.  But  the  duality  remains,  and 
must  be  carefully  preserved  in  thinking.  The  chief  practical  reason  for 
respecting  it  is  that  many  years  may  elapse  between  the  recognition  of  an 
invariant.,  experimentally  replicable  pattern  (including  its  proof  as  a 
pattern),  on  the  one  hand,  and  its  successful  interpretation  by  a  correctly 
named  and  conceived  source  trait  on  the  other.  During  this  period  in 
limbo,  it  is  important  to  preserve  the  pattern  with  a  label  which  is  as  far 
as  possible  descriptive  rather  than  interpretive.  For  the  downfall  of 

Personality  Theory  from  Quantitative  Research       273 

"faculty  psychology"  was  brought  about,  not  by  any  fallacy  in  the 
notion  of  a  faculty,  but  by  the  fact  that  the  faculties  were  allowed  to 
form  themselves  merely  on  the  patterns  of  existing  words.  Incidentally, 
the  odium  which  science  properly  attaches  to  this  verbal  vice  (even 
though  the  vice  is  now  driven  out  of  personality  research  where  it  most 
flourished),  still  attaches  itself  to  some  concepts  in  learning  theory  and 
comparative  physiological  psychology.  Nevertheless,  although  factor 
analysis  from  the  beginning  seeks  the  real  evidence  of  functional  co- 
variation, instead  of  unconsciously  accepting  the  false  unity  of  words, 
yet  the  premature  attachment  of  interpretive  labels  to  factors  may 
prejudice  real  freedom  of  thought  and  experiment.  It  is  for  this  reason, 
and  to  facilitate  work  on  the  establishment  of  factor  patterns  per  se3 
that  the  present  writer  has  suggested  a  Universal  Index,  -with  a  number 
for  each  pattern  believed  matched  over  at  least  three  independent 
studies  [26,  28].  Some  of  the  factors  believed  established  will  be  discussed 
in  the  following  section. 


Every  substantial  science  has  its  taxonomy.  Each  passes  through 
a  phase  in  which  greatest  activity  is  directed  to  producing  order  and 
stability  of  nomenclature,  before  its  more  comprehensive  theories — at 
least,  genuine  comprehensive  theories — can  hope  to  emerge.  So  in  per- 
sonality study,  before  "findings"  can  be  discussed  in  terms  of  purely 
psychological  concepts  and  laws,  some  statistical  and  methodological 
points  have  still  to  be  clarified  concerning  the  classification  and  ordering 
of  factor  patterns  per  se.  It  is  usual  to  speak  of  ability  factors  and  tem- 
perament factors,  of  general  and  specific  factors,  of  behavior  factors  and 
questionnaire  factors,  of  first-  and  second-order  factprs,  and  so  on.  How 
correct  is  it  to  use  these  categories,  and  on  what  are  they  founded? 
Perhaps  four  questions  will  get  to  the  heart  of  these  problems : 

1.  What  is  the  relation  of  a  factor  founded  on  behavioral  phenomena 
to  one  founded  on  introspective,  questionnaire  response? 

2.  How  do  we  know  that  the  factor  dimensions  we  obtain  span  the 
whole  personality,  or  some  given  domain  of  it? 

3.  How  do  we  know  when  a  factor  belongs  to  one  modality  or 
region,  e.g.,  that  it  is  an  ability  factor  rather  than  a  motivation  factor? 

4.  If  there  are  first-  and  second-order  factors,  how  do  we  know  at 
which  level  we  are  operating  in  a  given  case? 

The  first  two  questions  need  simultaneous  discussion.  The  question 
of  whether  a  factor  is  truly  general,  i.e.,  whether  it  spans  the  whole 
domain  of  human  behavior,  involves  also  asking  whether  experiment  has 


yet  covered  all  human  behavior.  To  ask  how  we  know  that  a  factor  is 
"general"  is  in  a  sense  as  ridiculous  as  asking,  "How  do  we  know  when 
the  science  of  physics  is  finished?"  But  consider  the  question,  "How  do 
geographers  know  when  all  new  land  has  been  found?"  and  it  will  be- 
come apparent  that  there  may,  nevertheless,  be  possibilities  of  progres- 
sively detailed  exploration  within  a  definite,  finite  area. 

Development  of  an  acceptable  notion  of  a  total,  definitive  area  of 
personality  manifestation  would  have  considerable  appeal  in  relation  to 
several  theoretical  problems  in  structured  (factor)  measurement.  In  the 
first  place  there  are  greater  difficulties  in  attempting  to  integrate  a  piece- 
meal, step-by-step  exploration  of  different  areas  of  variables  (as  is 
feasible  in  most  other  scientific  areas),  compared  with  those  encountered 
in  an  approach  attempting  to  "block  in"  the  main  perspectives  from  a 
total  realm  fixed  from  the  beginning.  But  a  "total  realm  of  phenomena" 
requires  the  concept  of  a  "population  of  variables,"  with  sampling 
properties  similar  to  that  of  a  population  of  persons.  (In  terms  of  R  and 
Q  techniques,  or  any  other  pair  of  transpose  techniques,  the  persons  and 
the  variables  have,  of  course,  just  such  a  reciprocal,  equivalent  relation- 
ship.) Variables  consist  of  stimulus-response  pairs,  so  in  principle  the 
possibility  exists  of  defining  a  total  population  of  stimuli,  response  habits, 
or  linkages  of  these,  within  a  given  culture  pattern.  This  special,  but 
basic,  issue  is  discussed  more  fully  in  an  appendix  to  the  present  article. 

Before  a  solution  is  suggested  here,  however,  it  behooves  us  to  note 
that  the  stimulus-response  behavior  of  the  human  organism  is  observed  in 
three  distinct  media.  It  may  be  observed  as  behavior,  embedded  in  the 
actual  life  situations,  in  which  case  we  will  call  it  the  life  record  medium, 
or  L  data.  Or  it  may  be  observed  as  introspective  responses  made  to  a 
questionnaire.  This  we  shall  call  Q  data.  A  good  deal  of  subtle  reasoning 
could  be  followed  up  about  Q  data,  but  the  main  point  is  that  they 
really  consist  of  two  distinct  kinds  of  data,  with  different  properties, 
according  to  how  the  questionnaire  is  scored.  If,  on  the  one  hand,  we 
accept  the  common  meaning  of  words,  i.e.,  accept  the  answers  as  fact 
about  the  individual's  consciousness  and  behavior,  we  shall  call  it  Q' 
data.  It  yields  "mental  interiors"  [14]  and  is  susceptible  to  no  reliability 
coefficient  between  two  equivalent,  but  different,  observers.  On  the 
other  hand,  if  the  answers  are  accepted  only  as  a  form  of  behavior, 
i.e.,  if  when  S  responds  "I  am  shy"  we  do  not  take  it  as  evidence  of 
shyness,  but  only  proof  that  S  so  responds,  we  shall  call  it  Q  data.  Such 
data  belongs  with  the  other  test  approach — objective  tests — and  so,  if  we 
abandon  the  introspective,  Q'  data,  we  have  essentially  only  two  media, 
the  life  situations  and  the  test  situations. 

Now  the  notion  of  a  population  of  variables  must  rest  on  the  life 
situations,  for  tests  can  be  multiplied  according  to  whim.  Lacking  re- 

Personality  Theory  from  Quantitative  Research       275 

sources  to  make  a  cultural  time-sampling  of  human  behavior  around 
the  clock,  the  present  writer  suggested  using  language  as  a  mirror  con- 
densing this  behavior.  It  was  assumed  that  the  dictionary  must,  by  the 
twentieth  century,  have  stabilized  the  number  of  symbols  required  to 
refer  to  all  aspects  of  human  behavior  of  interest  to  man.  This  symbol 
collection  is  called  the  personality  sphere,  envisaged  as  a  finite  but  un- 
bounded set  of  symbols,  which  can  be  represented  as  points  in  hyper- 
space  bearing  a  spatial  relation  to  one  another  which  is  some  function 
of  their  meaning  relationship. 

As  a  broad  strategy  of  research  it  has  been  advocated  [14,  28]  that 
source  traits  should  first  be  found  within  a  stratified  sample  of  variables 
based  on  this  form  of  L  data  (behavior  in  situ,  rated  with  special  pre- 
cautions). L  data  should  have  primacy  because  (1)  according  to  the 
above  argument,  we  can  be  reasonably  certain  of  covering  the  principal 
dimensions  thereby,  (2)  the  factors  will  appear  clothed  in  terms  already 
familiar  to  us  (in  everyday  and  clinical  language),  so  that  interpretive 
hypotheses  may  be  readily  reached,  and  (3)  the  construction  of  tests 
objectively  to  measure  these  primary  factors  will  then  be  guided  by  these 
hypotheses  and  will  no  longer  be  at  the  mercy  of  disproportionate  multi- 
plication of  test  behavior  merely  in  some  test-convenient  areas.  Enough 
of  factor  research  has  followed  this  strategy  to  permit  development  along 
the  lines  indicated. 

The  question  of  whether  a  factor  is  an  ability  or  a  dynamic  or  a 
temperamental  trait  hats  usually  been  confidently  decided,  among  most 
psychologists,  by  common  sense — until  those  numerous  borderline  cases 
arose  which  proved  common  sense  inadequate.  A  special  analysis  of  this 
problem  [14]  has  suggested  that  there  are  in  fact  three  possible  modalities 
of  factors,  as  implied  above,  though  any  given  variable  in  general 
expresses  in  varying  degrees  all  three  modalities.  A  variable  (trait  ele- 
ment) is  defined  as  dynamic  in  proportion  to  the  degree  that  the  mean 
score  (for  a  population)  changes  in  response  to  changes  in  the  incentive. 
When  the  score  is  in  an  "irrelevance  range"  of  immunity  to  changes  in 
incentive,  and  thus  becomes  sensitive  to  changes  in  complexity  of  the 
situation,  the  test  becomes  mainly  an  ability  test.  A  measure  which  is 
insensitive  to  both  changes  in  complexity  and  changes  in  incentive  is  de- 
fined as  a  temperament  measure.  Changes  in  the  situation  which  are  not 
changes  in  incentive  are  changes  in  complexity.  For  logical  completeness 
this  system  of  definitions  now  requires  an  independent  definition  of  an  in- 
centive. This  can  be  achieved  by  longitudinal  analysis  of  behavior,  de- 
fining a  goal  by  consummately  responses  at  which  a  train  of  behavior 
is  found  to  cease.  However,  there  are  complexities  in  the  modality  ques- 
tion which  require  such  space  for  discussion  that  the  reader  must  be 
referred  to  the  original  [14]  statement  of  theorems. 


Although  operational  definitions  of  the  three  varieties  of  modality  can 
thus  be  obtained,  to  supplant  the  rough  hunches  of  psychological 
common  sense,  yet  it  may  be  asked  whether  the  modality  classification 
of  factors — on  variables — has  any  value  except  as  an  academic  exercise. 
The  answer  to  this  would  seem  to  be  that  the  properties  of  abilities, 
temperament  traits,  and  dynamic  traits  differ  in  so  many  ways  that  there 
is  real  predictive  convenience  in  having  factors  as  far  as  possible  repre- 
senting purely  one  modality  or  another.  Now  the  factors  that  we  usually 
get  from  random  and  unassorted  variables  and  samples  will  normally 
be  w holistic  factors  [14-],  i.e.,  extending  across  modalities.  To  take  an 
even  broader  example,  in  a  group  of  children  ranging  from  five  to  ten 
years  where  physical  and  mental  variables  are  factored  together,  we 
might  get  a  single  growth  factor,  covering  both  intelligence  and  physical 
size.  On  the  other  hand,  in  most  factor  analytic  studies  we  actually  get 
conditional  factors,  i.e.,  factors  restricted  to  a  particular  modality,  be- 
cause the  variable  sample  is  suitably  restricted  and  all  the  variables  are 
in  any  case  presented  with  certain  conditions  retained  in  constancy.  For 
example,  in  ability  measures  a  high  and  sufficient  motivation  is  normally 
maintained  throughout,  whereas,  in  motivation  measures,  intellectual 
complications  as  such  are  implicitly  eliminated.  Consequently,  in  what 
follows  we  shall  generally  deal  with  (a)  ability,  (fc)  "general  personality 
and  temperament"  factors,  and  (c]  purely  motivational  factors,  wherever 
"conditional"  experiment  has  been  attempted.  However,  it  has  to  be 
admitted  that  complete  modality  separation  and  clarity  has  not  yet 
been  reached,  either  theoretically  or  in  the  findings. 

Our  fourth  question  dealt  with  the  general  nature  of  factors,  par- 
ticularly the  thorny  issue  of  determining  whether  a  factor  is  of  first, 
second,  or  higher  order.  It  might  seem  sufficient  to  say  that  any  factor 
found  by  factoring  an  initial  collection  of  operationally  defined  variables 
is  a  first-order  factor,  and  that  any  obtained  by  factoring  the  resultant 
factors  is  a  second-order  factor,  and  so  on.  A  little  reflection  will  show 
that  although  this  should  suffice  generally,  it  may  fail.  By  factoring  30 
to  50  varied  ability  tests,  Thurstone  and  others  [57,  90],  have  found 
about  a  dozen  primary  abilities.  The  simple  structure  shows  these  to  be 
oblique  in  relation  to  one  another.  When  the  correlation  matrix  among 
these  primary  factors  is  then  factored,  one  or  more  second-order  factors 
appear,  and  that  which  is  most  general  to  all  the  abilities  is  considered 
to  be  Spearman's  general  intelligence  factor.  It  has  been  shown  that  its 
loadings  directly  in  the  tests  are  the  same  as  those  obtained  for  Spear- 
man's general  factor,  "g"  [28].  Thus  general  ability  can  be  obtained 
either  as  a  first-order  (primary)  factor  or  as  a  second-order  factor,  though 
in  the  first  case  we  have  to  take  special  precautions  (tetrad  differences 
made  to  equal  zero)  in  choosing  the  variables  from  which  we  shall  work. 

Personality  Theory  from  Quantitative  Research       277 

Parenthetically  one  should  note  that  about  half  the  writers  on  second- 
order  factors  have  them  wrongly  conceived  as  factors  obtained  from  the 
correlations  of  either  (1)  the  reference  vectors,  when  indeed  a  cor- 
rected inverse  of  this  reference  vector  matrix  is  what  actually  has  to  be 
used,  or  (2)  the  factors  as  literally  experimentally  measured,  by  some 
battery  of  constituent  subtests.  Owing  to  the  immense  labor  and  skill 
required  accurately  to  determine  the  exact  hyperplane  angles  in  ( 1 )  as 
well  as  the  need  to  take  a  mean  of  several  studies,  no  data  fit  for  a 
second-order  analysis  of  personality  factors  have  been  available  until 
quite  recently  [28],  though  the  present  writer  must  confess  to  a  pre- 
mature attempt  to  determine  second-order  structure  in  1947. 

The  second-order  factors  of  general  anxiety,  extraversion,  etc.,  re- 
cently found  in  personality  will  be  mentioned  in  the  ensuing  brief  survey 
of  experimental  findings,  but  the  important  point  for  the  present  method- 
ological and  taxonomic  discussion  is  that  in  these,  as  with  the  older 
established  general  ability  factor,  it  has  happened  experimentally  that 
the  same  factor  has,  in  different  settings,  been  picked  up  both  as  a  first- 
order  and  a  second-order  factor.  Indeed  it  is  possible  to  see  theoretically 
that  this  could  happen.  For  example,  if  we  started  a  supposed  first-order 
ability  factorization  and  happened  to  use  as  variables  very  pure  measures 
of  Thurstone's  primaries,  and  only  one  measure  of  each,  our  factors  im- 
mediately obtained  would  be  factors  generally  encountered  only  as 
second-order  factors.  In  personality  most  measures  are  not  so  pure — 
the  proposed  measure  for  factor  A  contains  also  some  B,  C,  etc.  Con- 
sequently, if  there  are  enough  variables,  the  factoring  of  these  supposed 
first-order  factors  still  yields  first-order  factors,  as  was  found  in  Lovell's 
and  Thurstone's  reanalysis  of  Guilford's  highly  intercorrelating  question- 
naire measures  of  factors.  But  factorings  of  the  Sixteen  Personality  Factor 
Questionnaire,  with  its  relatively  pure  first-order  factor  measures,  have 
yielded  second-order  factors  immediately  [28]. 

Without  laboring  what  may  seem  an  unduly  technical  point,  let  us 
take  it  that  both  theoretically  and  practically  we  know  that,  with  rather 
unusual  circumstances,  it  is  possible  to  "go  right  through  the  floor"  di- 
rectly from  variables  to  second-order  factors  without  intervening  pri- 
maries. Consequently  one  cannot  infallibly  tell  on  which  floor  one  has 
landed  by  the  merely  operational  definition  that  first-order  factors  are 
what  you  get  from  test  variables  and  second-order  factors  are  what  you 
get  from  factoring  factors.  Other  things  being  equal,  the  danger  of  con- 
fusion through  going  directly  to  second-order  factors,  unknowingly,  from 
variables  is  much  greater  when  variables  are  very  diverse  in  nature 
and  chosen  sparsely  from  a  wide  area.  This  argument  that  factor  order 
is  related  to  density  of  variables  can  be  extended,  as  far  as  one  can  see,  to 
third-  and  higher-order  factors.  An  acceptable  taxonomy  and  classifica- 


tion  of  factors  according  to  order,  and  some  other  factor  concepts,  would 
therefore  be  assisted  by  an  operational  concept  of  density  of  variable 
sampling.  [For  more  adequate  examination  of  the  assumptions  in  vari- 
able density  the  reader  must  be  referre